fix(async-context-compression): resolve race condition, update role to assistant, bump to v1.1.3

Harden async context compression against new DB session alias and runtime edge cases

.agent/workflows/plugin-development.md

@@ -25,6 +25,10 @@ Every plugin **MUST** have bilingual versions for both code and documentation:
 - **Valves**: Use `pydantic` for configuration.
 - **Database**: Re-use `open_webui.internal.db` shared connection.
 - **User Context**: Use `_get_user_context` helper method.
+- **Chat API**: For message updates, follow the "OpenWebUI Chat API 更新规范" in `.github/copilot-instructions.md`.
+  - Use Event API for immediate UI updates
+  - Use Chat Persistence API for database storage
+  - Always update both `messages[]` and `history.messages`
 
 ### Commit Messages
 - **Language**: **English ONLY**. Do not use Chinese in commit messages.
@@ -35,8 +39,8 @@ Every plugin **MUST** have bilingual versions for both code and documentation:
 When adding or updating a plugin, you **MUST** update the following documentation files to maintain consistency:
 
 ### Plugin Directory
-- `README.md`: Update version, description, and usage.
-- `README_CN.md`: Update version, description, and usage.
+- `README.md`: Update version, description, and usage. **Explicitly describe new features in a prominent position at the beginning.**
+- `README_CN.md`: Update version, description, and usage. **Explicitly describe new features in a prominent position at the beginning.**
 
 ### Global Documentation (`docs/`)
 - **Index Pages**:
@@ -55,9 +59,13 @@ When adding or updating a plugin, you **MUST** update the following documentatio
 Reference: `.github/workflows/release.yml`
 
 ### Version Bumping
-- **Rule**: Any change to plugin logic **MUST** be accompanied by a version bump in the docstring.
+- **Rule**: Version bump is required **ONLY when the user explicitly requests a release**. Regular code changes do NOT require version bumps.
 - **Format**: Semantic Versioning (e.g., `1.0.0` -> `1.0.1`).
-- **Consistency**: Update version in **ALL** locations:
+- **When to Bump**: Only update the version when:
+  - User says "发布" / "release" / "bump version"
+  - User explicitly asks to prepare for release
+- **Agent Initiative**: After completing significant changes (new features, bug fixes, or multiple code modifications), the agent **SHOULD proactively ask** the user if they want to release a new version. If confirmed, update all version-related files.
+- **Consistency**: When bumping, update version in **ALL** locations:
   1. English Code (`.py`)
   2. Chinese Code (`.py`)
   3. English README (`README.md`)
@@ -74,6 +82,11 @@ Reference: `.github/workflows/release.yml`
     - Generates release notes based on changes.
     - Creates a GitHub Release tag (e.g., `v2024.01.01-1`).
     - Uploads individual `.py` files of **changed plugins only** as assets.
+4.  **Market Publishing**:
+    - Workflow: `.github/workflows/publish_plugin.yml`
+    - Trigger: Release published.
+    - Action: Automatically updates the plugin code and metadata on OpenWebUI.com using `scripts/publish_plugin.py`.
+    - Requirement: `OPENWEBUI_API_KEY` secret must be set.
 
 ### Pull Request Check
 - Workflow: `.github/workflows/plugin-version-check.yml`
@@ -91,3 +104,9 @@ Before committing:
 - [ ] `docs/` index and detail pages are updated?
 - [ ] Root `README.md` is updated?
 - [ ] All version numbers match exactly?
+
+## 5. Git Operations (Agent Rules)
+
+Strictly follow the rules defined in `.github/copilot-instructions.md` → **Git Operations (Agent Rules)** section.
+
+

.github/copilot-instructions.md

@@ -13,13 +13,13 @@ This document defines the standard conventions and best practices for OpenWebUI
 每个插件必须提供两个版本：
 
 1. **英文版本**: `plugin_name.py` - 英文界面、提示词和注释
-2. **中文版本**: `plugin_name_cn.py` 或 `插件中文名.py` - 中文界面、提示词和注释
+2. **中文版本**: `plugin_name_cn.py` - 中文界面、提示词和注释
 
 示例：
 ```
 plugins/actions/export_to_docx/
 ├── export_to_word.py      # English version
-├── 导出为Word.py           # Chinese version
+├── export_to_word_cn.py    # Chinese version
 ├── README.md               # English documentation
 └── README_CN.md            # Chinese documentation
 ```
@@ -31,13 +31,75 @@ plugins/actions/export_to_docx/
 - `README.md` - English documentation
 - `README_CN.md` - 中文文档
 
-README 文件应包含以下内容：
-- 功能描述 / Feature description
-- 配置参数及默认值 / Configuration parameters with defaults
-- 安装和设置说明 / Installation and setup instructions
-- 使用示例 / Usage examples
-- 故障排除指南 / Troubleshooting guide
-- 版本和作者信息 / Version and author information
+### README 结构规范 (README Structure Standard)
+
+所有插件 README 必须遵循以下统一结构顺序：
+
+1.  **标题 (Title)**: 插件名称，带 Emoji 图标
+2.  **元数据 (Metadata)**: 作者、版本、项目链接 (一行显示)
+    - 格式: `**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** x.x.x | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)`
+    - **注意**: Author 和 Project 为固定值，仅需更新 Version 版本号
+3.  **描述 (Description)**: 一句话功能介绍
+4.  **最新更新 (What's New)**: **必须**放在描述之后，显著展示最新版本的变更点 (仅展示最近 3 次更新)
+5.  **核心特性 (Key Features)**: 使用 Emoji + 粗体标题 + 描述格式
+6.  **使用方法 (How to Use)**: 按步骤说明
+7.  **配置参数 (Configuration/Valves)**: 使用表格格式，包含参数名、默认值、描述
+8.  **其他 (Others)**: 支持的模板类型、语法示例、故障排除等
+
+完整示例 (Full Example):
+
+```markdown
+# 📊 Smart Plugin
+
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.0.0 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
+
+A one-sentence description of this plugin.
+
+## 🔥 What's New in v1.0.0
+
+- ✨ **Feature Name**: Brief description of the feature.
+- 🔧 **Configuration Change**: What changed in settings.
+- 🐛 **Bug Fix**: What was fixed.
+
+## ✨ Key Features
+
+- 🚀 **Feature A**: Description of feature A.
+- 🎨 **Feature B**: Description of feature B.
+- 📥 **Feature C**: Description of feature C.
+
+## 🚀 How to Use
+
+1. **Install**: Search for "Plugin Name" in the Open WebUI Community and install.
+2. **Trigger**: Enter your text in the chat, then click the **Action Button**.
+3. **Result**: View the generated result.
+
+## ⚙️ Configuration (Valves)
+
+| Parameter | Default | Description |
+| :--- | :--- | :--- |
+| **Show Status (SHOW_STATUS)** | `True` | Whether to show status updates. |
+| **Model ID (MODEL_ID)** | `Empty` | LLM model for processing. |
+| **Output Mode (OUTPUT_MODE)** | `image` | `image` for static, `html` for interactive. |
+
+## 🛠️ Supported Types (Optional)
+
+| Category | Type Name | Use Case |
+| :--- | :--- | :--- |
+| **Category A** | `type-a`, `type-b` | Use case description |
+
+## 📝 Advanced Example (Optional)
+
+\`\`\`syntax
+example code or syntax here
+\`\`\`
+```
+
+### 文档内容要求 (Content Requirements)
+
+- **新增功能**: 必须在 "What's New" 章节中明确列出，使用 Emoji + 粗体标题格式 (仅保留最近 3 个版本的更新记录)。
+- **双语**: 必须提供 `README.md` (英文) 和 `README_CN.md` (中文)。
+- **表格对齐**: 配置参数表格使用左对齐 `:---`。
+- **Emoji 规范**: 标题使用合适的 Emoji 增强可读性。
 
 ### 官方文档 (Official Documentation)
 
@@ -91,33 +153,7 @@ icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0i...（完整的 Base64 编
 
 ---
 
-## 👤 作者和许可证信息 (Author and License)
-
-所有 README 文件和主要文档必须包含以下统一信息：
-
-```markdown
-## Author
-
-Fu-Jie  
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## License
-
-MIT License
-```
-
-中文版本：
-
-```markdown
-## 作者
-
-Fu-Jie  
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## 许可证
-
-MIT License
-```
+(Author info is now part of the top metadata section, see "README Structure Standard" above)
 
 ---
 
@@ -224,7 +260,46 @@ async def _emit_notification(
 
 ## 📋 日志规范 (Logging Standard)
 
-- **禁止使用** `print()` 语句
+### 1. 前端控制台调试 (Frontend Console Debugging) - **优先推荐 (Preferred)**
+
+对于需要实时查看数据流、排查 UI 交互或内容变更的场景，**优先使用**前端控制台日志。这种方式可以直接在浏览器 DevTools (F12) 中查看，无需访问服务端日志。
+
+**实现方式**: 通过 `__event_emitter__` 发送 `type: "execute"` 事件执行 JS 代码。
+
+```python
+import json
+
+async def _emit_debug_log(self, __event_emitter__, title: str, data: dict):
+    """在浏览器控制台打印结构化调试日志"""
+    if not self.valves.show_debug_log or not __event_emitter__:
+        return
+
+    try:
+        js_code = f"""
+            (async function() {{
+                console.group("🛠️ {title}");
+                console.log({json.dumps(data, ensure_ascii=False)});
+                console.groupEnd();
+            }})();
+        """
+        
+        await __event_emitter__({
+            "type": "execute",
+            "data": {"code": js_code}
+        })
+    except Exception as e:
+        print(f"Error emitting debug log: {e}")
+```
+
+**配置要求**:
+- 在 `Valves` 中添加 `show_debug_log: bool` 开关，默认关闭。
+- 仅在开关开启时发送日志。
+
+### 2. 服务端日志 (Server-side Logging)
+
+用于记录系统级错误、异常堆栈或无需前端感知的后台任务。
+
+- **禁止使用** `print()` 语句 (除非用于简单的脚本调试)
 - 必须使用 Python 标准库 `logging`
 
 ```python
@@ -505,7 +580,164 @@ Base = declarative_base()
 
 ---
 
-## 🔧 代码规范 (Code Style)
+## 📂 文件存储访问规范 (File Storage Access)
+
+OpenWebUI 支持多种文件存储后端（本地磁盘、S3/MinIO 对象存储等）。插件在访问用户上传的文件或生成的图片时，必须实现多级回退机制以兼容所有存储配置。
+
+### 存储类型检测 (Storage Type Detection)
+
+通过 `Files.get_file_by_id()` 获取的文件对象，其 `path` 属性决定了存储位置：
+
+| Path 格式 | 存储类型 | 访问方式 |
+|-----------|----------|----------|
+| `s3://bucket/key` | S3/MinIO 对象存储 | boto3 直连或 API 回调 |
+| `/app/backend/data/...` | Docker 卷存储 | 本地文件系统读取 |
+| `./uploads/...` | 本地相对路径 | 本地文件系统读取 |
+| `gs://bucket/key` | Google Cloud Storage | API 回调 |
+
+### 多级回退机制 (Multi-level Fallback)
+
+推荐实现以下优先级的文件获取策略：
+
+```python
+def _get_file_content(self, file_id: str, max_bytes: int) -> Optional[bytes]:
+    """获取文件内容，支持多种存储后端"""
+    file_obj = Files.get_file_by_id(file_id)
+    if not file_obj:
+        return None
+
+    # 1️⃣ 数据库直接存储 (小文件)
+    data_field = getattr(file_obj, "data", None)
+    if isinstance(data_field, dict):
+        if "bytes" in data_field:
+            return data_field["bytes"]
+        if "base64" in data_field:
+            return base64.b64decode(data_field["base64"])
+
+    # 2️⃣ S3 直连 (对象存储 - 最快)
+    s3_path = getattr(file_obj, "path", None)
+    if isinstance(s3_path, str) and s3_path.startswith("s3://"):
+        data = self._read_from_s3(s3_path, max_bytes)
+        if data:
+            return data
+
+    # 3️⃣ 本地文件系统 (磁盘存储)
+    for attr in ("path", "file_path"):
+        path = getattr(file_obj, attr, None)
+        if path and not path.startswith(("s3://", "gs://", "http")):
+            # 尝试多个常见路径
+            for base in ["", "./data", "/app/backend/data"]:
+                full_path = Path(base) / path if base else Path(path)
+                if full_path.exists():
+                    return full_path.read_bytes()[:max_bytes]
+
+    # 4️⃣ 公共 URL 下载
+    url = getattr(file_obj, "url", None)
+    if url and url.startswith("http"):
+        return self._download_from_url(url, max_bytes)
+
+    # 5️⃣ 内部 API 回调 (通用兜底方案)
+    if self._api_base_url:
+        api_url = f"{self._api_base_url}/api/v1/files/{file_id}/content"
+        return self._download_from_api(api_url, self._api_token, max_bytes)
+
+    return None
+```
+
+### S3 直连实现 (S3 Direct Access)
+
+当检测到 `s3://` 路径时，使用 `boto3` 直接访问对象存储，读取以下环境变量：
+
+| 环境变量 | 说明 | 示例 |
+|----------|------|------|
+| `S3_ENDPOINT_URL` | S3 兼容服务端点 | `https://minio.example.com` |
+| `S3_ACCESS_KEY_ID` | 访问密钥 ID | `minioadmin` |
+| `S3_SECRET_ACCESS_KEY` | 访问密钥 | `minioadmin` |
+| `S3_ADDRESSING_STYLE` | 寻址样式 | `auto`, `path`, `virtual` |
+
+```python
+# S3 直连示例
+import boto3
+from botocore.config import Config as BotoConfig
+import os
+
+def _read_from_s3(self, s3_path: str, max_bytes: int) -> Optional[bytes]:
+    """从 S3 直接读取文件 (比 API 回调更快)"""
+    if not s3_path.startswith("s3://"):
+        return None
+
+    # 解析 s3://bucket/key
+    parts = s3_path[5:].split("/", 1)
+    bucket, key = parts[0], parts[1]
+
+    # 从环境变量读取配置
+    endpoint = os.environ.get("S3_ENDPOINT_URL")
+    access_key = os.environ.get("S3_ACCESS_KEY_ID")
+    secret_key = os.environ.get("S3_SECRET_ACCESS_KEY")
+    
+    if not all([endpoint, access_key, secret_key]):
+        return None  # 回退到 API 方式
+
+    s3_client = boto3.client(
+        "s3",
+        endpoint_url=endpoint,
+        aws_access_key_id=access_key,
+        aws_secret_access_key=secret_key,
+        config=BotoConfig(s3={"addressing_style": os.environ.get("S3_ADDRESSING_STYLE", "auto")})
+    )
+    
+    response = s3_client.get_object(Bucket=bucket, Key=key)
+    return response["Body"].read(max_bytes)
+```
+
+### API 回调实现 (API Fallback)
+
+当其他方式失败时，通过 OpenWebUI 内部 API 获取文件：
+
+```python
+def _download_from_api(self, api_url: str, token: str, max_bytes: int) -> Optional[bytes]:
+    """通过 OpenWebUI API 获取文件内容"""
+    import urllib.request
+    
+    headers = {"User-Agent": "OpenWebUI-Plugin"}
+    if token:
+        headers["Authorization"] = token
+
+    req = urllib.request.Request(api_url, headers=headers)
+    with urllib.request.urlopen(req, timeout=15) as response:
+        if 200 <= response.status < 300:
+            return response.read(max_bytes)
+    return None
+```
+
+### 获取 API 上下文 (API Context Extraction)
+
+在 `action()` 方法中捕获请求上下文，用于 API 回调：
+
+```python
+async def action(self, body: dict, __request__=None, ...):
+    # 从请求对象获取 API 凭证
+    if __request__:
+        self._api_token = __request__.headers.get("Authorization")
+        self._api_base_url = str(__request__.base_url).rstrip("/")
+    else:
+        # 从环境变量获取端口作为备用
+        port = os.environ.get("PORT") or "8080"
+        self._api_base_url = f"http://localhost:{port}"
+        self._api_token = None
+```
+
+### 性能对比 (Performance Comparison)
+
+| 方式 | 网络跳数 | 适用场景 |
+|------|----------|----------|
+| S3 直连 | 1 (插件 → S3) | 对象存储，最快 |
+| 本地文件 | 0 | 磁盘存储，最快 |
+| API 回调 | 2 (插件 → OpenWebUI → S3/磁盘) | 通用兜底 |
+
+### 参考实现 (Reference Implementation)
+
+- `plugins/actions/export_to_docx/export_to_word.py` - `_image_bytes_from_owui_file_id` 方法
 
 ### Python 规范
 
@@ -796,10 +1028,371 @@ For iframe plugins to access parent document theme information, users need to co
 - [ ] 使用 logging 而非 print
 - [ ] 测试双语界面
 - [ ] **一致性检查 (Consistency Check)**:
-    - [ ] 更新 `README.md` 插件列表
-    - [ ] 更新 `README_CN.md` 插件列表
-    - [ ] 更新/创建 `docs/` 下的对应文档
-    - [ ] 确保文档版本号与代码一致
+
+---
+
+## 🚀 高级开发模式 (Advanced Development Patterns)
+
+### 混合服务端-客户端生成 (Hybrid Server-Client Generation)
+
+对于需要复杂前端渲染（如 Mermaid 图表、ECharts）但最终生成文件（如 DOCX、PDF）的场景，建议采用混合模式：
+
+1.  **服务端 (Python)**：
+    *   处理文本解析、Markdown 转换、文档结构构建。
+    *   为复杂组件生成**占位符**（如带有特定 ID 或元数据的图片/文本块）。
+    *   将半成品文件（如 Base64 编码的 ZIP/DOCX）发送给前端。
+
+2.  **客户端 (JavaScript)**：
+    *   在浏览器中加载半成品文件（使用 JSZip 等库）。
+    *   利用浏览器能力渲染复杂组件（如 `mermaid.render`）。
+    *   将渲染结果（SVG/PNG）回填到占位符位置。
+    *   触发最终文件的下载。
+
+**优势**：
+*   无需在服务端安装 Headless Browser（如 Puppeteer），降低部署复杂度。
+*   利用用户浏览器的计算能力。
+*   支持动态、交互式内容的静态化导出。
+
+### 原生 Word 公式支持 (Native Word Math Support)
+
+对于需要生成高质量数学公式的 Word 文档，推荐使用 `latex2mathml` + `mathml2omml` 组合：
+
+1.  **LaTeX -> MathML**: 使用 `latex2mathml` 将 LaTeX 字符串转换为标准 MathML。
+2.  **MathML -> OMML**: 使用 `mathml2omml` 将 MathML 转换为 Office Math Markup Language (OMML)。
+3.  **插入 Word**: 将 OMML XML 插入到 `python-docx` 的段落中。
+
+```python
+# 示例代码
+from latex2mathml.converter import convert as latex2mathml
+from mathml2omml import convert as mathml2omml
+
+def add_math(paragraph, latex_str):
+    mathml = latex2mathml(latex_str)
+    omml = mathml2omml(mathml)
+    # ... 插入 OMML 到 paragraph._element ...
+```
+
+### JS 渲染并嵌入 Markdown (JS Render to Markdown)
+
+对于需要复杂前端渲染（如 AntV 图表、Mermaid 图表、ECharts）但希望结果**持久化为纯 Markdown 格式**的场景，推荐使用 Data URL 嵌入模式：
+
+#### 工作流程
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│                    Plugin Workflow                          │
+├─────────────────────────────────────────────────────────────┤
+│  1. Python Action                                            │
+│     ├── 分析消息内容                                          │
+│     ├── 调用 LLM 生成结构化数据（可选）                        │
+│     └── 通过 __event_call__ 发送 JS 代码到前端                │
+├─────────────────────────────────────────────────────────────┤
+│  2. Browser JS (via __event_call__)                          │
+│     ├── 动态加载可视化库（如 AntV、Mermaid）                   │
+│     ├── 离屏渲染 SVG/Canvas                                   │
+│     ├── 使用 toDataURL() 导出 Base64 Data URL                 │
+│     └── 通过 REST API 更新消息内容                            │
+├─────────────────────────────────────────────────────────────┤
+│  3. Markdown 渲染                                            │
+│     └── 显示 ![描述](data:image/svg+xml;base64,...)          │
+└─────────────────────────────────────────────────────────────┘
+```
+
+#### 核心实现代码
+
+**Python 端（发送 JS 执行）：**
+
+```python
+async def action(self, body, __event_call__, __metadata__, ...):
+    chat_id = self._extract_chat_id(body, __metadata__)
+    message_id = self._extract_message_id(body, __metadata__)
+    
+    # 生成 JS 代码
+    js_code = self._generate_js_code(
+        chat_id=chat_id,
+        message_id=message_id,
+        data=processed_data,  # 可视化所需数据
+    )
+    
+    # 执行 JS
+    if __event_call__:
+        await __event_call__({
+            "type": "execute",
+            "data": {"code": js_code}
+        })
+```
+
+**JavaScript 端（渲染并回写）：**
+
+```javascript
+(async function() {
+    // 1. 动态加载可视化库
+    if (typeof VisualizationLib === 'undefined') {
+        await new Promise((resolve, reject) => {
+            const script = document.createElement('script');
+            script.src = 'https://cdn.example.com/lib.min.js';
+            script.onload = resolve;
+            script.onerror = reject;
+            document.head.appendChild(script);
+        });
+    }
+    
+    // 2. 创建离屏容器
+    const container = document.createElement('div');
+    container.style.cssText = 'position:absolute;left:-9999px;';
+    document.body.appendChild(container);
+    
+    // 3. 渲染可视化
+    const instance = new VisualizationLib({ container, ... });
+    instance.render(data);
+    
+    // 4. 导出为 Data URL
+    const dataUrl = await instance.toDataURL({ type: 'svg', embedResources: true });
+    // 或手动转换 SVG:
+    // const svgData = new XMLSerializer().serializeToString(svgElement);
+    // const base64 = btoa(unescape(encodeURIComponent(svgData)));
+    // const dataUrl = "data:image/svg+xml;base64," + base64;
+    
+    // 5. 清理
+    instance.destroy();
+    document.body.removeChild(container);
+    
+    // 6. 生成 Markdown 图片
+    const markdownImage = `![描述](${dataUrl})`;
+    
+    // 7. 通过 API 更新消息
+    const token = localStorage.getItem("token");
+    await fetch(`/api/v1/chats/${chatId}/messages/${messageId}/event`, {
+        method: "POST",
+        headers: {
+            "Content-Type": "application/json",
+            "Authorization": `Bearer ${token}`
+        },
+        body: JSON.stringify({
+            type: "chat:message",
+            data: { content: originalContent + "\n\n" + markdownImage }
+        })
+    });
+})();
+```
+
+#### 优势
+
+- **纯 Markdown 输出**：结果是标准的 Markdown 图片语法，无需 HTML 代码块
+- **高效存储**：图片上传至 `/api/v1/files`，避免 Base64 字符串膨胀聊天记录
+- **持久化**：通过 API 回写，消息重新加载后图片仍然存在
+- **跨平台**：任何支持 Markdown 图片的客户端都能显示
+- **无服务端渲染依赖**：利用用户浏览器的渲染能力
+
+#### 与 HTML 注入模式对比
+
+| 特性 | HTML 注入 (`\`\`\`html`) | JS 渲染 + Markdown 图片 |
+|------|-------------------------|------------------------|
+| 输出格式 | HTML 代码块 | Markdown 图片 |
+| 交互性 | ✅ 支持按钮、动画 | ❌ 静态图片 |
+| 外部依赖 | 需要加载 JS 库 | 依赖 `/api/v1/files` 存储 |
+| 持久化 | 依赖浏览器渲染 | ✅ 永久可见 |
+| 文件导出 | 需特殊处理 | ✅ 直接导出 |
+| 适用场景 | 交互式内容 | 信息图、图表快照 |
+
+#### 参考实现
+
+- `plugins/actions/js-render-poc/infographic_markdown.py` - AntV Infographic 生成并嵌入
+- `plugins/actions/js-render-poc/js_render_poc.py` - 基础概念验证
+
+### OpenWebUI Chat API 更新规范 (Chat API Update Specification)
+
+当插件需要修改消息内容并持久化到数据库时，必须遵循 OpenWebUI 的 Backend-Controlled API 流程。
+
+When a plugin needs to modify message content and persist it to the database, follow OpenWebUI's Backend-Controlled API flow.
+
+#### 核心概念 (Core Concepts)
+
+1. **Event API** (`/api/v1/chats/{chatId}/messages/{messageId}/event`)
+   - 用于**即时更新前端显示**，用户无需刷新页面
+   - 是可选的，部分版本可能不支持
+   - 仅影响当前会话的 UI，不持久化
+
+2. **Chat Persistence API** (`/api/v1/chats/{chatId}`)
+   - 用于**持久化到数据库**，确保刷新页面后数据仍存在
+   - 必须同时更新 `messages[]` 数组和 `history.messages` 对象
+   - 是消息持久化的唯一可靠方式
+
+#### 数据结构 (Data Structure)
+
+OpenWebUI 的 Chat 对象包含两个关键位置存储消息内容：
+
+```javascript
+{
+  "chat": {
+    "id": "chat-uuid",
+    "title": "Chat Title",
+    "messages": [                              // 1️⃣ 消息数组
+      { "id": "msg-1", "role": "user", "content": "..." },
+      { "id": "msg-2", "role": "assistant", "content": "..." }
+    ],
+    "history": {
+      "current_id": "msg-2",
+      "messages": {                            // 2️⃣ 消息索引对象
+        "msg-1": { "id": "msg-1", "role": "user", "content": "..." },
+        "msg-2": { "id": "msg-2", "role": "assistant", "content": "..." }
+      }
+    }
+  }
+}
+```
+
+> **重要**：修改消息时，**必须同时更新两个位置**，否则可能导致数据不一致。
+
+#### 标准实现流程 (Standard Implementation)
+
+```javascript
+(async function() {
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    const token = localStorage.getItem("token");
+    
+    // 1️⃣ 获取当前 Chat 数据
+    const getResponse = await fetch(`/api/v1/chats/${chatId}`, {
+        method: "GET",
+        headers: { "Authorization": `Bearer ${token}` }
+    });
+    const chatData = await getResponse.json();
+    
+    // 2️⃣ 使用 map 遍历 messages，只修改目标消息
+    let newContent = "";
+    const updatedMessages = chatData.chat.messages.map(m => {
+        if (m.id === messageId) {
+            const originalContent = m.content || "";
+            newContent = originalContent + "\n\n" + newMarkdown;
+            
+            // 3️⃣ 同时更新 history.messages 中对应的消息
+            if (chatData.chat.history && chatData.chat.history.messages) {
+                if (chatData.chat.history.messages[messageId]) {
+                    chatData.chat.history.messages[messageId].content = newContent;
+                }
+            }
+            
+            // 4️⃣ 保留消息的其他属性，只修改 content
+            return { ...m, content: newContent };
+        }
+        return m;  // 其他消息原样返回
+    });
+    
+    // 5️⃣ 通过 Event API 即时更新前端（可选）
+    try {
+        await fetch(`/api/v1/chats/${chatId}/messages/${messageId}/event`, {
+            method: "POST",
+            headers: {
+                "Content-Type": "application/json",
+                "Authorization": `Bearer ${token}`
+            },
+            body: JSON.stringify({
+                type: "chat:message",
+                data: { content: newContent }
+            })
+        });
+    } catch (e) {
+        // Event API 是可选的，继续执行持久化
+        console.log("Event API not available, continuing...");
+    }
+    
+    // 6️⃣ 持久化到数据库（必须）
+    const updatePayload = {
+        chat: {
+            ...chatData.chat,      // 保留所有原有属性
+            messages: updatedMessages
+            // history 已在上面原地修改
+        }
+    };
+    
+    await fetch(`/api/v1/chats/${chatId}`, {
+        method: "POST",
+        headers: {
+            "Content-Type": "application/json",
+            "Authorization": `Bearer ${token}`
+        },
+        body: JSON.stringify(updatePayload)
+    });
+})();
+```
+
+#### 最佳实践 (Best Practices)
+
+1. **保留原有结构**：使用展开运算符 `...chatData.chat` 和 `...m` 确保不丢失任何原有属性
+2. **双位置更新**：必须同时更新 `messages[]` 和 `history.messages[id]`
+3. **错误处理**：Event API 调用应包裹在 try-catch 中，失败时继续持久化
+4. **重试机制**：对持久化 API 实现重试逻辑，提高可靠性
+
+```javascript
+// 带重试的请求函数
+const fetchWithRetry = async (url, options, retries = 3) => {
+    for (let i = 0; i < retries; i++) {
+        try {
+            const response = await fetch(url, options);
+            if (response.ok) return response;
+            if (i < retries - 1) {
+                await new Promise(r => setTimeout(r, 1000 * (i + 1)));  // 指数退避
+            }
+        } catch (e) {
+            if (i === retries - 1) throw e;
+            await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+        }
+    }
+    return null;
+};
+```
+
+5. **禁止使用的 API**：不要使用 `/api/v1/chats/{chatId}/share` 作为持久化备用方案，该 API 用于分享功能，不是更新功能
+
+#### 提取 Chat ID 和 Message ID (Extracting IDs)
+
+```python
+def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+    """从 body 或 metadata 中提取 chat_id"""
+    if isinstance(body, dict):
+        chat_id = body.get("chat_id")
+        if isinstance(chat_id, str) and chat_id.strip():
+            return chat_id.strip()
+        
+        body_metadata = body.get("metadata", {})
+        if isinstance(body_metadata, dict):
+            chat_id = body_metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+    
+    if isinstance(metadata, dict):
+        chat_id = metadata.get("chat_id")
+        if isinstance(chat_id, str) and chat_id.strip():
+            return chat_id.strip()
+    
+    return ""
+
+def _extract_message_id(self, body: dict, metadata: Optional[dict]) -> str:
+    """从 body 或 metadata 中提取 message_id"""
+    if isinstance(body, dict):
+        message_id = body.get("id")
+        if isinstance(message_id, str) and message_id.strip():
+            return message_id.strip()
+        
+        body_metadata = body.get("metadata", {})
+        if isinstance(body_metadata, dict):
+            message_id = body_metadata.get("message_id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+    
+    if isinstance(metadata, dict):
+        message_id = metadata.get("message_id")
+        if isinstance(message_id, str) and message_id.strip():
+            return message_id.strip()
+    
+    return ""
+```
+
+#### 参考实现
+
+- `plugins/actions/smart-mind-map/smart_mind_map.py` - 思维导图图片模式实现
+- 官方文档: [Backend-Controlled, UI-Compatible API Flow](https://docs.openwebui.com/tutorials/tips/backend-controlled-ui-compatible-api-flow)
 
 ---
 
@@ -831,13 +1424,35 @@ For iframe plugins to access parent document theme information, users need to co
 
 ### 发布前必须完成 (Pre-release Requirements)
 
+> [!IMPORTANT]
+> 版本号**仅在用户明确要求发布时**才需要更新。日常代码更改**无需**更新版本号。
+
+**触发版本更新的关键词**：
+- 用户说 "发布"、"release"、"bump version"
+- 用户明确要求准备发布
+
+**Agent 主动询问发布 (Agent-Initiated Release Prompt)**：
+
+当 Agent 完成以下类型的更改后，**应主动询问**用户是否需要发布新版本：
+
+| 更改类型 | 示例 | 是否询问发布 |
+|---------|------|-------------|
+| 新功能 | 新增导出格式、新的配置选项 | ✅ 询问 |
+| 重要 Bug 修复 | 修复导致崩溃或数据丢失的问题 | ✅ 询问 |
+| 累积多次更改 | 同一插件在会话中被修改 >= 3 次 | ✅ 询问 |
+| 小优化 | 代码清理、格式符号处理 | ❌ 不询问 |
+| 文档更新 | 只改 README、注释 | ❌ 不询问 |
+
+如果用户确认发布，Agent 需要更新所有版本相关的文件（代码、README、docs 等）。
+
+**发布时需要完成**：
 1. ✅ **更新版本号** - 修改插件文档字符串中的 `version` 字段
 2. ✅ **中英文版本同步** - 确保两个版本的版本号一致
 
 ```python
 """
 title: My Plugin
-version: 0.2.0  # <- 必须更新这里！
+version: 0.2.0  # <- 发布时更新这里！
 ...
 """
 ```
@@ -982,3 +1597,83 @@ Follow the [Conventional Commits](https://www.conventionalcommits.org/) specific
 ❌ **Bad:**
 - `新增导出PDF插件` (Chinese is not allowed)
 - `update code` (Too vague)
+
+---
+
+## 🤖 Git Operations (Agent Rules)
+
+**重要规则 (CRITICAL RULES FOR AI AGENTS)**:
+
+AI Agent（如 Copilot、Gemini、Claude 等）在执行 Git 操作时必须遵守以下规则：
+
+| 操作 (Operation) | 允许 (Allowed) | 说明 (Description) |
+|-----------------|---------------|---------------------|
+| 创建功能分支 | ✅ 允许 | `git checkout -b feature/xxx` |
+| 推送到功能分支 | ✅ 允许 | `git push origin feature/xxx` |
+| 直接推送到 main | ❌ 禁止 | `git push origin main` 需要用户手动执行 |
+| 合并到 main | ❌ 禁止 | 任何合并操作需要用户明确批准 |
+| Rebase 到 main | ❌ 禁止 | 任何 rebase 操作需要用户明确批准 |
+
+**规则详解 (Rule Details)**:
+
+1. **Feature Branches Allowed**: Agent **可以**创建新的功能分支并推送到远程仓库
+2. **No Direct Push to Main**: Agent **禁止**直接推送任何更改到 `main` 分支
+3. **No Auto-Merge**: Agent **禁止**在未经用户明确批准的情况下合并任何分支到 `main`
+4. **User Approval Required**: 任何影响 `main` 分支的操作（push、merge、rebase）都需要用户明确批准
+
+> [!CAUTION]
+> 违反上述规则可能导致代码库不稳定或触发意外的 CI/CD 流程。Agent 应始终在功能分支上工作，并让用户决定何时合并到主分支。
+
+---
+
+## ⏳ 长时间运行任务通知 (Long-running Task Notifications)
+
+如果一个前台任务（Foreground Task）的运行时间预计超过 **3秒**，必须实现用户通知机制，以避免用户感到困惑。
+
+**要求 (Requirements):**
+
+1. **初始通知 (Initial Notification)**: 任务开始时**立即**发送第一条通知，告知用户正在处理中（例如：“正在使用 AI 生成中...”）。
+2. **周期性通知 (Periodic Notification)**: 之后每隔 **5秒** 发送一次通知，告知用户任务仍在运行中。
+3. **完成清理 (Cleanup)**: 任务完成后，应自动取消通知任务。
+
+**代码示例 (Code Example):**
+
+```python
+import asyncio
+
+async def long_running_task_with_notification(self, event_emitter, ...):
+    # 定义实际任务
+    async def actual_task():
+        # ... 执行耗时操作 ...
+        return result
+
+    # 定义通知任务
+    async def notification_task():
+        # 立即发送首次通知
+        if event_emitter:
+            await self._send_notification(event_emitter, "info", "正在使用 AI 生成中...")
+        
+        # 之后每5秒通知一次
+        while True:
+            await asyncio.sleep(5)
+            if event_emitter:
+                await self._send_notification(event_emitter, "info", "仍在处理中，请耐心等待...")
+
+    # 并发运行任务
+    task_future = asyncio.ensure_future(actual_task())
+    notify_future = asyncio.ensure_future(notification_task())
+
+    # 等待任务完成
+    done, pending = await asyncio.wait(
+        [task_future, notify_future], 
+        return_when=asyncio.FIRST_COMPLETED
+    )
+
+    # 取消通知任务
+    if not notify_future.done():
+        notify_future.cancel()
+
+    # 获取结果
+    if task_future in done:
+        return task_future.result()
+```

.github/workflows/community-stats.yml

@@ -0,0 +1,78 @@
+# OpenWebUI 社区统计报告自动生成
+# 只在统计数据变化时 commit，避免频繁提交
+
+name: Community Stats
+
+on:
+  # 每小时整点运行
+  schedule:
+    - cron: '0 * * * *'
+  # 手动触发
+  workflow_dispatch:
+
+permissions:
+  contents: write
+
+jobs:
+  update-stats:
+    runs-on: ubuntu-latest
+    
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          token: ${{ secrets.GITHUB_TOKEN }}
+      
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+      
+      - name: Install dependencies
+        run: |
+          pip install requests python-dotenv
+      
+      - name: Get previous stats
+        id: prev_stats
+        run: |
+          # 获取当前的 points 用于比较
+          if [ -f docs/community-stats.json ]; then
+            OLD_POINTS=$(jq -r '.user.total_points' docs/community-stats.json 2>/dev/null || echo "0")
+            echo "old_points=$OLD_POINTS" >> $GITHUB_OUTPUT
+          else
+            echo "old_points=0" >> $GITHUB_OUTPUT
+          fi
+      
+      - name: Generate stats report
+        env:
+          OPENWEBUI_API_KEY: ${{ secrets.OPENWEBUI_API_KEY }}
+          OPENWEBUI_USER_ID: ${{ secrets.OPENWEBUI_USER_ID }}
+        run: |
+          python scripts/openwebui_stats.py
+      
+      - name: Check for significant changes
+        id: check_changes
+        run: |
+          # 获取新的 points
+          NEW_POINTS=$(jq -r '.user.total_points' docs/community-stats.json 2>/dev/null || echo "0")
+          
+          echo "📊 Previous points: ${{ steps.prev_stats.outputs.old_points }}"
+          echo "📊 Current points: $NEW_POINTS"
+          
+          # 只在 points 变化时才 commit
+          if [ "$NEW_POINTS" != "${{ steps.prev_stats.outputs.old_points }}" ]; then
+            echo "changed=true" >> $GITHUB_OUTPUT
+            echo "✅ Points changed (${{ steps.prev_stats.outputs.old_points }} → $NEW_POINTS), will commit"
+          else
+            echo "changed=false" >> $GITHUB_OUTPUT
+            echo "⏭️ Points unchanged, skipping commit"
+          fi
+      
+      - name: Commit and push changes
+        if: steps.check_changes.outputs.changed == 'true'
+        run: |
+          git config --local user.email "github-actions[bot]@users.noreply.github.com"
+          git config --local user.name "github-actions[bot]"
+          git add docs/community-stats.zh.md docs/community-stats.md docs/community-stats.json README.md README_CN.md
+          git diff --staged --quiet || git commit -m "chore: update community stats $(date +'%Y-%m-%d')"
+          git push

.github/workflows/publish_new_plugin.yml

@@ -0,0 +1,68 @@
+name: Publish New Plugin
+
+on:
+  workflow_dispatch:
+    inputs:
+      plugin_dir:
+        description: 'Plugin directory (e.g., plugins/actions/deep-dive)'
+        required: true
+        type: string
+      dry_run:
+        description: 'Dry run mode (preview only)'
+        required: false
+        type: boolean
+        default: false
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.x'
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install requests
+
+      - name: Validate plugin directory
+        run: |
+          if [ ! -d "${{ github.event.inputs.plugin_dir }}" ]; then
+            echo "❌ Error: Directory '${{ github.event.inputs.plugin_dir }}' does not exist"
+            exit 1
+          fi
+          echo "✅ Found plugin directory: ${{ github.event.inputs.plugin_dir }}"
+          ls -la "${{ github.event.inputs.plugin_dir }}"
+
+      - name: Publish Plugin
+        env:
+          OPENWEBUI_API_KEY: ${{ secrets.OPENWEBUI_API_KEY }}
+        run: |
+          if [ "${{ github.event.inputs.dry_run }}" = "true" ]; then
+            echo "🔍 Dry run mode - previewing..."
+            python scripts/publish_plugin.py --new "${{ github.event.inputs.plugin_dir }}" --dry-run
+          else
+            echo "🚀 Publishing plugin..."
+            python scripts/publish_plugin.py --new "${{ github.event.inputs.plugin_dir }}"
+          fi
+
+      - name: Commit changes (if ID was added)
+        if: ${{ github.event.inputs.dry_run != 'true' }}
+        run: |
+          git config user.name "github-actions[bot]"
+          git config user.email "github-actions[bot]@users.noreply.github.com"
+          
+          # Check if there are changes to commit
+          if git diff --quiet; then
+            echo "No changes to commit"
+          else
+            git add "${{ github.event.inputs.plugin_dir }}"
+            git commit -m "feat: add openwebui_id to ${{ github.event.inputs.plugin_dir }}"
+            git push
+            echo "✅ Committed and pushed openwebui_id changes"
+          fi

.github/workflows/publish_plugin.yml

@@ -0,0 +1,33 @@
+name: Publish Plugins to OpenWebUI Market
+
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - 'plugins/**/*.py'
+  release:
+    types: [published]
+  workflow_dispatch:
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.x'
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install requests
+
+      - name: Publish Plugins
+        env:
+          OPENWEBUI_API_KEY: ${{ secrets.OPENWEBUI_API_KEY }}
+        run: python scripts/publish_plugin.py

.github/workflows/release.yml

@@ -54,6 +54,9 @@ permissions:
 jobs:
   check-changes:
     runs-on: ubuntu-latest
+    env:
+      LANG: en_US.UTF-8
+      LC_ALL: en_US.UTF-8
     outputs:
       has_changes: ${{ steps.detect.outputs.has_changes }}
       changed_plugins: ${{ steps.detect.outputs.changed_plugins }}
@@ -65,6 +68,12 @@ jobs:
         with:
           fetch-depth: 0
 
+      - name: Configure Git
+        run: |
+          git config --global core.quotepath false
+          git config --global i18n.commitencoding utf-8
+          git config --global i18n.logoutputencoding utf-8
+
       - name: Set up Python
         uses: actions/setup-python@v5
         with:
@@ -131,6 +140,7 @@ jobs:
             
             echo "changed_plugins<<EOF" >> $GITHUB_OUTPUT
             cat changed_files.txt >> $GITHUB_OUTPUT
+            echo "" >> $GITHUB_OUTPUT
             echo "EOF" >> $GITHUB_OUTPUT
           fi
           
@@ -138,6 +148,7 @@ jobs:
           {
             echo 'release_notes<<EOF'
             cat changes.md
+            echo ""
             echo 'EOF'
           } >> $GITHUB_OUTPUT
 
@@ -145,6 +156,10 @@ jobs:
     needs: check-changes
     if: needs.check-changes.outputs.has_changes == 'true' || github.event_name == 'workflow_dispatch' || startsWith(github.ref, 'refs/tags/v')
     runs-on: ubuntu-latest
+    env:
+      LANG: en_US.UTF-8
+      LC_ALL: en_US.UTF-8
+
     
     steps:
       - name: Checkout repository
@@ -152,6 +167,12 @@ jobs:
         with:
           fetch-depth: 0
 
+      - name: Configure Git
+        run: |
+          git config --global core.quotepath false
+          git config --global i18n.commitencoding utf-8
+          git config --global i18n.logoutputencoding utf-8
+
       - name: Set up Python
         uses: actions/setup-python@v5
         with:
@@ -159,14 +180,34 @@ jobs:
 
       - name: Determine version
         id: version
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: |
           if [ "${{ github.event_name }}" = "workflow_dispatch" ] && [ -n "${{ github.event.inputs.version }}" ]; then
             VERSION="${{ github.event.inputs.version }}"
           elif [[ "${{ github.ref }}" == refs/tags/v* ]]; then
             VERSION="${GITHUB_REF#refs/tags/}"
           else
-            # Auto-generate version based on date and run number
-            VERSION="v$(date +'%Y.%m.%d')-${{ github.run_number }}"
+            # Auto-generate version based on date and daily release count
+            TODAY=$(date +'%Y.%m.%d')
+            TODAY_PREFIX="v${TODAY}-"
+            
+            # Count existing releases with today's date prefix
+            # grep -c returns 1 if count is 0, so we use || true to avoid script failure
+            EXISTING_COUNT=$(gh release list --limit 100 2>/dev/null | grep -c "^${TODAY_PREFIX}" || true)
+            
+            # Clean up output (handle potential newlines or fallback issues)
+            EXISTING_COUNT=$(echo "$EXISTING_COUNT" | tr -cd '0-9')
+            if [ -z "$EXISTING_COUNT" ]; then EXISTING_COUNT=0; fi
+            
+            NEXT_NUM=$((EXISTING_COUNT + 1))
+            
+            VERSION="${TODAY_PREFIX}${NEXT_NUM}"
+            
+            # Final fallback to ensure VERSION is never empty
+            if [ -z "$VERSION" ]; then
+              VERSION="v$(date +'%Y.%m.%d-%H%M%S')"
+            fi
           fi
           echo "version=$VERSION" >> $GITHUB_OUTPUT
           echo "Release version: $VERSION"
@@ -205,6 +246,17 @@ jobs:
           echo "=== Collected Files ==="
           find release_plugins -name "*.py" -type f | head -20
 
+      - name: Debug Filenames
+        run: |
+          python3 -c "import sys; print(f'Filesystem encoding: {sys.getfilesystemencoding()}')"
+          ls -R release_plugins
+
+      - name: Upload Debug Artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: debug-plugins
+          path: release_plugins/
+
       - name: Get commit messages
         id: commits
         if: github.event_name == 'push'
@@ -220,8 +272,9 @@ jobs:
           {
             echo 'commits<<EOF'
             echo "$COMMITS"
+            echo ""
             echo 'EOF'
-          } >> $GITHUB_OUTPUT
+          } >> "$GITHUB_OUTPUT"
 
       - name: Generate release notes
         id: notes
@@ -292,19 +345,51 @@ jobs:
           echo "=== Release Notes ==="
           cat release_notes.md
 
+      - name: Create Git Tag
+        run: |
+          VERSION="${{ steps.version.outputs.version }}"
+          
+          if [ -z "$VERSION" ]; then
+            echo "Error: Version is empty!"
+            exit 1
+          fi
+          
+          if ! git rev-parse "$VERSION" >/dev/null 2>&1; then
+            echo "Creating tag $VERSION"
+            git tag "$VERSION"
+            git push origin "$VERSION"
+          else
+            echo "Tag $VERSION already exists"
+          fi
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
       - name: Create GitHub Release
         uses: softprops/action-gh-release@v2
         with:
           tag_name: ${{ steps.version.outputs.version }}
+          target_commitish: ${{ github.sha }}
           name: ${{ github.event.inputs.release_title || steps.version.outputs.version }}
           body_path: release_notes.md
           prerelease: ${{ github.event.inputs.prerelease || false }}
+          make_latest: true
           files: |
             plugin_versions.json
-            release_plugins/**/*.py
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
+      - name: Upload Release Assets
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          # Check if there are any .py files to upload
+          if [ -d release_plugins ] && [ -n "$(find release_plugins -type f -name '*.py' 2>/dev/null)" ]; then
+            echo "Uploading plugin files..."
+            find release_plugins -type f -name "*.py" -print0 | xargs -0 gh release upload ${{ steps.version.outputs.version }} --clobber
+          else
+            echo "No plugin files to upload. Skipping asset upload."
+          fi
+
       - name: Summary
         run: |
           echo "## 🚀 Release Created Successfully!" >> $GITHUB_STEP_SUMMARY

CONTRIBUTING.md

@@ -1,87 +1,16 @@
-# 贡献指南 (Contributing Guide)
+# Contributing Guide
 
-感谢你对 **OpenWebUI Extras** 感兴趣！我们非常欢迎社区贡献更多的插件、提示词和创意。
+Thank you for your interest in **OpenWebUI Extras**!
 
-## 🤝 如何贡献
+## 🚀 How to Contribute
 
-### 1. 分享提示词 (Prompts)
+1. **Fork** this repository.
+2. **Add/Modify** the plugin file in the `plugins/` directory.
+3. **Submit PR**: We will review and merge it.
 
-如果你有一个好用的提示词：
-1. 在 `prompts/` 目录下找到合适的分类（如 `coding/`, `writing/`）。如果没有合适的，可以新建一个文件夹。
-2. 创建一个新的 `.md` 或 `.json` 文件。
-3. 提交 Pull Request (PR)。
+## 💡 Important
 
-### 2. 开发插件 (Plugins)
+- Ensure your plugin includes complete metadata (title, author, version, description).
+- If updating an existing plugin, please **increment the version number** (e.g., `0.1.0` -> `0.1.1`) to trigger the auto-update.
 
-如果你开发了一个新的 OpenWebUI 插件 (Function/Tool)：
-1. 确保你的插件代码包含完整的元数据（Frontmatter）：
-   ```python
-   """
-   title: 插件名称
-   author: 你的名字
-   version: 0.1.0
-   description: 简短描述插件的功能
-   """
-   ```
-2. 将插件文件放入 `plugins/` 目录下的合适位置：
-   - `plugins/actions/`: 用于添加按钮或修改消息的 Action 插件。
-   - `plugins/filters/`: 用于拦截请求或响应的 Filter 插件。
-   - `plugins/pipes/`: 用于自定义模型或 API 的 Pipe 插件。
-   - `plugins/tools/`: 用于 LLM 调用的 Tool 插件。
-3. 建议在 `docs/` 下添加一个简单的使用说明。
-
-### 3. 改进文档
-
-如果你发现文档有错误或可以改进的地方，直接提交 PR 即可。
-
-## 🛠️ 开发规范
-
-- **代码风格**：Python 代码请遵循 PEP 8 规范。
-- **注释**：关键逻辑请添加注释，方便他人理解。
-- **测试**：提交前请在本地 OpenWebUI 环境中测试通过。
-
-## 📝 提交 PR
-
-1. Fork 本仓库。
-2. 创建一个新的分支 (`git checkout -b feature/AmazingFeature`)。
-3. 提交你的修改 (`git commit -m 'Add some AmazingFeature'`)。
-4. 推送到分支 (`git push origin feature/AmazingFeature`)。
-5. 开启一个 Pull Request。
-
-## 📦 版本更新与发布
-
-当你更新插件时，请遵循以下流程：
-
-### 1. 更新版本号
-
-在插件文件的 docstring 中更新版本号（遵循[语义化版本](https://semver.org/lang/zh-CN/)）：
-
-```python
-"""
-title: 我的插件
-version: 0.2.0  # 更新此处
-...
-"""
-```
-
-### 2. 更新更新日志
-
-在 `CHANGELOG.md` 的 `[Unreleased]` 部分添加你的更改：
-
-```markdown
-### Added / 新增
-- 新功能描述
-
-### Fixed / 修复
-- Bug 修复描述
-```
-
-### 3. 发布流程
-
-维护者会通过以下方式发布新版本：
-- 手动触发 GitHub Actions 中的 "Plugin Release" 工作流
-- 或创建版本标签 (`v*`)
-
-详细说明请参阅 [发布工作流文档](docs/release-workflow.zh.md)。
-
-再次感谢你的贡献！🚀
+Thank you! 🚀

CONTRIBUTING_CN.md

@@ -0,0 +1,16 @@
+# 贡献指南
+
+感谢你对 **OpenWebUI Extras** 感兴趣！
+
+## 🚀 贡献流程
+
+1. **Fork** 本仓库。
+2. **修改/添加** `plugins/` 目录下的插件文件。
+3. **提交 PR**: 我们会尽快审核并合并。
+
+## 💡 注意事项
+
+- 请确保插件包含完整的元数据（title, author, version, description）。
+- 如果是更新已有插件，请记得**增加版本号**（如 `0.1.0` -> `0.1.1`），这样系统会自动同步更新。
+
+再次感谢你的贡献！🚀

README.md

@@ -4,7 +4,34 @@ English | [中文](./README_CN.md)
 
 A collection of enhancements, plugins, and prompts for [OpenWebUI](https://github.com/open-webui/open-webui), developed and curated for personal use to extend functionality and improve experience.
 
-[Contributing](./CONTRIBUTING.md)
+<!-- STATS_START -->
+## 📊 Community Stats
+
+> 🕐 Auto-updated: 2026-01-12 01:06
+
+| 👤 Author | 👥 Followers | ⭐ Points | 🏆 Contributions |
+|:---:|:---:|:---:|:---:|
+| [Fu-Jie](https://openwebui.com/u/Fu-Jie) | **82** | **86** | **22** |
+
+| 📝 Posts | ⬇️ Downloads | 👁️ Views | 👍 Upvotes | 💾 Saves |
+|:---:|:---:|:---:|:---:|:---:|
+| **14** | **1161** | **12713** | **76** | **73** |
+
+### 🔥 Top 6 Popular Plugins
+
+> 🕐 Auto-updated: 2026-01-12 01:06
+
+| Rank | Plugin | Downloads | Views | Updated |
+|:---:|------|:---:|:---:|:---:|
+| 🥇 | [Smart Mind Map](https://openwebui.com/posts/turn_any_text_into_beautiful_mind_maps_3094c59a) | 368 | 3352 | 2026-01-07 |
+| 🥈 | [Export to Excel](https://openwebui.com/posts/export_mulit_table_to_excel_244b8f9d) | 184 | 590 | 2026-01-07 |
+| 🥉 | [📊 Smart Infographic (AntV)](https://openwebui.com/posts/smart_infographic_ad6f0c7f) | 136 | 1482 | 2026-01-11 |
+| 4️⃣ | [Async Context Compression](https://openwebui.com/posts/async_context_compression_b1655bc8) | 136 | 1510 | 2026-01-11 |
+| 5️⃣ | [Flash Card](https://openwebui.com/posts/flash_card_65a2ea8f) | 100 | 1833 | 2026-01-07 |
+| 6️⃣ | [Export to Word (Enhanced)](https://openwebui.com/posts/export_to_word_enhanced_formatting_fca6a315) | 96 | 880 | 2026-01-07 |
+
+*See full stats in [Community Stats Report](./docs/community-stats.md)*
+<!-- STATS_END -->
 
 ## 📦 Project Contents
 
@@ -15,15 +42,18 @@ Located in the `plugins/` directory, containing Python-based enhancements:
 #### Actions
 - **Smart Mind Map** (`smart-mind-map`): Generates interactive mind maps from text.
 - **Smart Infographic** (`infographic`): Transforms text into professional infographics using AntV.
-- **Knowledge Card** (`knowledge-card`): Creates beautiful flashcards for learning.
+- **Flash Card** (`flash-card`): Quickly generates beautiful flashcards for learning.
+- **Deep Dive** (`deep-dive`): A comprehensive thinking lens that dives deep into any content.
 - **Export to Excel** (`export_to_excel`): Exports chat history to Excel files.
 - **Export to Word** (`export_to_docx`): Exports chat history to Word documents.
-- **Summary** (`summary`): Text summarization tool.
 
 #### Filters
 - **Async Context Compression** (`async-context-compression`): Optimizes token usage via context compression.
 - **Context Enhancement** (`context_enhancement_filter`): Enhances chat context.
 - **Gemini Manifold Companion** (`gemini_manifold_companion`): Companion filter for Gemini Manifold.
+- **Gemini Multimodal Filter** (`web_gemini_multimodel_filter`): Provides multimodal capabilities (PDF, Office, Video) for any model via Gemini.
+- **Markdown Normalizer** (`markdown_normalizer`): Fixes common Markdown formatting issues in LLM outputs.
+- **Multi-Model Context Merger** (`multi_model_context_merger`): Automatically merges and injects context from multiple model responses.
 
 
 #### Pipes
@@ -60,10 +90,16 @@ This project is a collection of resources and does not require a Python environm
 
 ### Using Plugins
 
-1. Browse the `/plugins` directory and download the plugin file (`.py`) you need.
-2. Go to OpenWebUI **Admin Panel** -> **Settings** -> **Plugins**.
-3. Click the upload button and select the `.py` file you just downloaded.
-4. Once uploaded, refresh the page to enable the plugin in your chat settings or toolbar.
+1. **Install from OpenWebUI Community (Recommended)**:
+   - Visit my profile: [Fu-Jie's Profile](https://openwebui.com/u/Fu-Jie)
+   - Browse the plugins and select the one you like.
+   - Click "Get" to import it directly into your OpenWebUI instance.
+
+2. **Manual Installation**:
+   - Browse the `/plugins` directory and download the plugin file (`.py`) you need.
+   - Go to OpenWebUI **Admin Panel** -> **Settings** -> **Plugins**.
+   - Click the upload button and select the `.py` file you just downloaded.
+   - Once uploaded, refresh the page to enable the plugin in your chat settings or toolbar.
 
 ### Contributing
 
@@ -71,3 +107,5 @@ If you have great prompts or plugins to share:
 1. Fork this repository.
 2. Add your files to the appropriate `prompts/` or `plugins/` directory.
 3. Submit a Pull Request.
+
+[Contributing](./CONTRIBUTING.md)

README_CN.md

@@ -2,23 +2,58 @@
 
 [English](./README.md) | 中文
 
-OpenWebUI 增强功能集合。包含个人开发与收集的### 🧩 插件 (Plugins)
+OpenWebUI 增强功能集合。包含个人开发与收集的插件、提示词等资源。
+
+<!-- STATS_START -->
+## 📊 社区统计
+
+> 🕐 自动更新于 2026-01-12 01:06
+
+| 👤 作者 | 👥 粉丝 | ⭐ 积分 | 🏆 贡献 |
+|:---:|:---:|:---:|:---:|
+| [Fu-Jie](https://openwebui.com/u/Fu-Jie) | **82** | **86** | **22** |
+
+| 📝 发布 | ⬇️ 下载 | 👁️ 浏览 | 👍 点赞 | 💾 收藏 |
+|:---:|:---:|:---:|:---:|:---:|
+| **14** | **1161** | **12713** | **76** | **73** |
+
+### 🔥 热门插件 Top 6
+
+> 🕐 自动更新于 2026-01-12 01:06
+
+| 排名 | 插件 | 下载 | 浏览 | 更新日期 |
+|:---:|------|:---:|:---:|:---:|
+| 🥇 | [Smart Mind Map](https://openwebui.com/posts/turn_any_text_into_beautiful_mind_maps_3094c59a) | 368 | 3352 | 2026-01-07 |
+| 🥈 | [Export to Excel](https://openwebui.com/posts/export_mulit_table_to_excel_244b8f9d) | 184 | 590 | 2026-01-07 |
+| 🥉 | [📊 Smart Infographic (AntV)](https://openwebui.com/posts/smart_infographic_ad6f0c7f) | 136 | 1482 | 2026-01-11 |
+| 4️⃣ | [Async Context Compression](https://openwebui.com/posts/async_context_compression_b1655bc8) | 136 | 1510 | 2026-01-11 |
+| 5️⃣ | [Flash Card](https://openwebui.com/posts/flash_card_65a2ea8f) | 100 | 1833 | 2026-01-07 |
+| 6️⃣ | [Export to Word (Enhanced)](https://openwebui.com/posts/export_to_word_enhanced_formatting_fca6a315) | 96 | 880 | 2026-01-07 |
+
+*完整统计请查看 [社区统计报告](./docs/community-stats.zh.md)*
+<!-- STATS_END -->
+
+## 📦 项目内容
+
+### 🧩 插件 (Plugins)
 
 位于 `plugins/` 目录，包含各类 Python 编写的功能增强插件：
 
 #### Actions (交互增强)
 - **Smart Mind Map** (`smart-mind-map`): 智能分析文本并生成交互式思维导图。
 - **Smart Infographic** (`infographic`): 基于 AntV 的智能信息图生成工具。
-- **Knowledge Card** (`knowledge-card`): 快速生成精美的学习记忆卡片。
+- **Flash Card** (`flash-card`): 快速生成精美的学习记忆卡片。
+- **Deep Dive** (`deep-dive`): 深度思考透镜，从背景、逻辑、洞察到行动路径的全方位分析。
 - **Export to Excel** (`export_to_excel`): 将对话内容导出为 Excel 文件。
 - **Export to Word** (`export_to_docx`): 将对话内容导出为 Word 文档。
-- **Summary** (`summary`): 文本摘要生成工具。
 
 #### Filters (消息处理)
 - **Async Context Compression** (`async-context-compression`): 异步上下文压缩，优化 Token 使用。
 - **Context Enhancement** (`context_enhancement_filter`): 上下文增强过滤器。
 - **Gemini Manifold Companion** (`gemini_manifold_companion`): Gemini Manifold 配套增强。
-
+- **Gemini Multimodal Filter** (`web_gemini_multimodel_filter`): 为任意模型提供多模态能力（PDF、Office、视频等），支持智能路由和字幕精修。
+- **Markdown Normalizer** (`markdown_normalizer`): 修复 LLM 输出中常见的 Markdown 格式问题。
+- **Multi-Model Context Merger** (`multi_model_context_merger`): 自动合并并注入多模型回答的上下文。
 
 #### Pipes (模型管道)
 - **Gemini Manifold** (`gemini_mainfold`): 集成 Gemini 模型的管道。
@@ -31,40 +66,10 @@ OpenWebUI 增强功能集合。包含个人开发与收集的### 🧩 插件 (Pl
 位于 `prompts/` 目录，包含精心调优的 System Prompts：
 
 - **Coding**: 编程辅助类提示词。
-- **Marketing**: 营销文案类提示词。(`/prompts/marketing`): 内容创作、品牌策划、市场分析相关的提示词
+- **Marketing**: 营销文案类提示词。
 
 每个提示词都独立保存为 Markdown 文件，可直接在 OpenWebUI 中使用。
 
-### 🔧 插件 (Plugins)
-
-{{ ... }}
-
-[贡献指南](./CONTRIBUTING.md) | [更新日志](./CHANGELOG.md)
-
-## 📦 项目内容
-
-### 🎯 提示词 (Prompts)
-
-位于 `/prompts` 目录，包含针对不同领域的优质提示词模板：
-
-- **编程类** (`/prompts/coding`): 代码生成、调试、优化相关的提示词
-- **营销类** (`/prompts/marketing`): 内容创作、品牌策划、市场分析相关的提示词
-
-每个提示词都独立保存为 Markdown 文件，可直接在 OpenWebUI 中使用。
-
-### 🔧 插件 (Plugins)
-
-位于 `/plugins` 目录，提供三种类型的插件扩展：
-
-- **过滤器 (Filters)** - 在用户输入发送给 LLM 前进行处理和优化
-  - 异步上下文压缩：智能压缩长上下文，优化 token 使用效率
-
-- **动作 (Actions)** - 自定义功能，从聊天中触发
-  - 思维导图生成：快速生成和导出思维导图
-
-- **管道 (Pipes)** - 对 LLM 响应进行处理和增强
-  - 各类响应处理和格式化插件
-
 ## 📖 开发文档
 
 位于 `docs/zh/` 目录：
@@ -87,10 +92,16 @@ OpenWebUI 增强功能集合。包含个人开发与收集的### 🧩 插件 (Pl
 
 ### 使用插件 (Plugins)
 
-1. 在 `/plugins` 目录中浏览并下载你需要的插件文件 (`.py`)。
-2. 打开 OpenWebUI 的 **管理员面板 (Admin Panel)** -> **设置 (Settings)** -> **插件 (Plugins)**。
-3. 点击上传按钮，选择刚才下载的 `.py` 文件。
-4. 上传成功后，刷新页面，你就可以在聊天设置或工具栏中启用该插件了。
+1. **从 OpenWebUI 社区安装 (推荐)**:
+   - 访问我的主页: [Fu-Jie's Profile](https://openwebui.com/u/Fu-Jie)
+   - 浏览插件列表，选择你喜欢的插件。
+   - 点击 "Get" 按钮，将其直接导入到你的 OpenWebUI 实例中。
+
+2. **手动安装**:
+   - 在 `/plugins` 目录中浏览并下载你需要的插件文件 (`.py`)。
+   - 打开 OpenWebUI 的 **管理员面板 (Admin Panel)** -> **设置 (Settings)** -> **插件 (Plugins)**。
+   - 点击上传按钮，选择刚才下载的 `.py` 文件。
+   - 上传成功后，刷新页面，你就可以在聊天设置或工具栏中启用该插件了。
 
 ### 贡献代码
 
@@ -98,3 +109,5 @@ OpenWebUI 增强功能集合。包含个人开发与收集的### 🧩 插件 (Pl
 1. Fork 本仓库。
 2. 将你的文件添加到对应的 `prompts/` 或 `plugins/` 目录。
 3. 提交 Pull Request。
+
+[贡献指南](./CONTRIBUTING_CN.md) | [更新日志](./CHANGELOG.md)

docs/community-stats.json

@@ -0,0 +1,252 @@
+{
+  "total_posts": 14,
+  "total_downloads": 1161,
+  "total_views": 12713,
+  "total_upvotes": 76,
+  "total_downvotes": 2,
+  "total_saves": 73,
+  "total_comments": 18,
+  "by_type": {
+    "action": 11,
+    "filter": 2,
+    "unknown": 1
+  },
+  "posts": [
+    {
+      "title": "Smart Mind Map",
+      "slug": "turn_any_text_into_beautiful_mind_maps_3094c59a",
+      "type": "action",
+      "version": "0.9.1",
+      "author": "Fu-Jie",
+      "description": "Intelligently analyzes text content and generates interactive mind maps to help users structure and visualize knowledge.",
+      "downloads": 368,
+      "views": 3352,
+      "upvotes": 11,
+      "saves": 22,
+      "comments": 11,
+      "created_at": "2025-12-30",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/turn_any_text_into_beautiful_mind_maps_3094c59a"
+    },
+    {
+      "title": "Export to Excel",
+      "slug": "export_mulit_table_to_excel_244b8f9d",
+      "type": "action",
+      "version": "0.3.7",
+      "author": "Fu-Jie",
+      "description": "Extracts tables from chat messages and exports them to Excel (.xlsx) files with smart formatting.",
+      "downloads": 184,
+      "views": 590,
+      "upvotes": 3,
+      "saves": 4,
+      "comments": 0,
+      "created_at": "2025-05-30",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/export_mulit_table_to_excel_244b8f9d"
+    },
+    {
+      "title": "📊 Smart Infographic (AntV)",
+      "slug": "smart_infographic_ad6f0c7f",
+      "type": "action",
+      "version": "1.4.9",
+      "author": "Fu-Jie",
+      "description": "AI-powered infographic generator based on AntV Infographic. Supports professional templates, auto-icon matching, and SVG/PNG downloads.",
+      "downloads": 136,
+      "views": 1482,
+      "upvotes": 8,
+      "saves": 9,
+      "comments": 2,
+      "created_at": "2025-12-28",
+      "updated_at": "2026-01-11",
+      "url": "https://openwebui.com/posts/smart_infographic_ad6f0c7f"
+    },
+    {
+      "title": "Async Context Compression",
+      "slug": "async_context_compression_b1655bc8",
+      "type": "filter",
+      "version": "1.1.2",
+      "author": "Fu-Jie",
+      "description": "Reduces token consumption in long conversations while maintaining coherence through intelligent summarization and message compression.",
+      "downloads": 136,
+      "views": 1510,
+      "upvotes": 6,
+      "saves": 10,
+      "comments": 0,
+      "created_at": "2025-11-08",
+      "updated_at": "2026-01-11",
+      "url": "https://openwebui.com/posts/async_context_compression_b1655bc8"
+    },
+    {
+      "title": "Flash Card",
+      "slug": "flash_card_65a2ea8f",
+      "type": "action",
+      "version": "0.2.4",
+      "author": "Fu-Jie",
+      "description": "Quickly generates beautiful flashcards from text, extracting key points and categories.",
+      "downloads": 100,
+      "views": 1833,
+      "upvotes": 8,
+      "saves": 6,
+      "comments": 2,
+      "created_at": "2025-12-30",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/flash_card_65a2ea8f"
+    },
+    {
+      "title": "Export to Word (Enhanced)",
+      "slug": "export_to_word_enhanced_formatting_fca6a315",
+      "type": "action",
+      "version": "0.4.3",
+      "author": "Fu-Jie",
+      "description": "Export current conversation from Markdown to Word (.docx) with Mermaid diagrams rendered client-side (Mermaid.js, SVG+PNG), LaTeX math, real hyperlinks, improved tables, syntax highlighting, and blockquote support.",
+      "downloads": 96,
+      "views": 880,
+      "upvotes": 6,
+      "saves": 9,
+      "comments": 0,
+      "created_at": "2026-01-03",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/export_to_word_enhanced_formatting_fca6a315"
+    },
+    {
+      "title": "📊 智能信息图 (AntV Infographic)",
+      "slug": "智能信息图_e04a48ff",
+      "type": "action",
+      "version": "1.4.9",
+      "author": "Fu-Jie",
+      "description": "基于 AntV Infographic 的智能信息图生成插件。支持多种专业模板，自动图标匹配，并提供 SVG/PNG 下载功能。",
+      "downloads": 35,
+      "views": 511,
+      "upvotes": 3,
+      "saves": 0,
+      "comments": 0,
+      "created_at": "2025-12-28",
+      "updated_at": "2026-01-11",
+      "url": "https://openwebui.com/posts/智能信息图_e04a48ff"
+    },
+    {
+      "title": "导出为 Word (增强版)",
+      "slug": "导出为_word_支持公式流程图表格和代码块_8a6306c0",
+      "type": "action",
+      "version": "0.4.3",
+      "author": "Fu-Jie",
+      "description": "将对话导出为 Word (.docx)，支持 Mermaid 图表 (客户端渲染 SVG+PNG)、LaTeX 数学公式、真实超链接、增强表格格式、代码高亮和引用块。",
+      "downloads": 35,
+      "views": 998,
+      "upvotes": 9,
+      "saves": 2,
+      "comments": 1,
+      "created_at": "2026-01-04",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/导出为_word_支持公式流程图表格和代码块_8a6306c0"
+    },
+    {
+      "title": "Deep Dive",
+      "slug": "deep_dive_c0b846e4",
+      "type": "action",
+      "version": "1.0.0",
+      "author": "Fu-Jie",
+      "description": "A comprehensive thinking lens that dives deep into any content - from context to logic, insights, and action paths.",
+      "downloads": 30,
+      "views": 307,
+      "upvotes": 3,
+      "saves": 4,
+      "comments": 0,
+      "created_at": "2026-01-08",
+      "updated_at": "2026-01-08",
+      "url": "https://openwebui.com/posts/deep_dive_c0b846e4"
+    },
+    {
+      "title": "思维导图",
+      "slug": "智能生成交互式思维导图帮助用户可视化知识_8d4b097b",
+      "type": "action",
+      "version": "0.9.1",
+      "author": "Fu-Jie",
+      "description": "智能分析文本内容,生成交互式思维导图,帮助用户结构化和可视化知识。",
+      "downloads": 18,
+      "views": 330,
+      "upvotes": 2,
+      "saves": 1,
+      "comments": 0,
+      "created_at": "2025-12-31",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/智能生成交互式思维导图帮助用户可视化知识_8d4b097b"
+    },
+    {
+      "title": "闪记卡 (Flash Card)",
+      "slug": "闪记卡生成插件_4a31eac3",
+      "type": "action",
+      "version": "0.2.4",
+      "author": "Fu-Jie",
+      "description": "快速将文本提炼为精美的学习记忆卡片，支持核心要点提取与分类。",
+      "downloads": 12,
+      "views": 361,
+      "upvotes": 4,
+      "saves": 1,
+      "comments": 0,
+      "created_at": "2025-12-30",
+      "updated_at": "2026-01-07",
+      "url": "https://openwebui.com/posts/闪记卡生成插件_4a31eac3"
+    },
+    {
+      "title": "异步上下文压缩",
+      "slug": "异步上下文压缩_5c0617cb",
+      "type": "filter",
+      "version": "1.1.2",
+      "author": "Fu-Jie",
+      "description": "通过智能摘要和消息压缩，降低长对话的 token 消耗，同时保持对话连贯性。",
+      "downloads": 8,
+      "views": 211,
+      "upvotes": 4,
+      "saves": 1,
+      "comments": 0,
+      "created_at": "2025-11-08",
+      "updated_at": "2026-01-11",
+      "url": "https://openwebui.com/posts/异步上下文压缩_5c0617cb"
+    },
+    {
+      "title": "精读",
+      "slug": "精读_99830b0f",
+      "type": "action",
+      "version": "1.0.0",
+      "author": "Fu-Jie",
+      "description": "全方位的思维透镜 —— 从背景全景到逻辑脉络，从深度洞察到行动路径。",
+      "downloads": 3,
+      "views": 112,
+      "upvotes": 2,
+      "saves": 1,
+      "comments": 0,
+      "created_at": "2026-01-08",
+      "updated_at": "2026-01-08",
+      "url": "https://openwebui.com/posts/精读_99830b0f"
+    },
+    {
+      "title": " 🛠️ Debug Open WebUI Plugins in Your Browser",
+      "slug": "debug_open_webui_plugins_in_your_browser_81bf7960",
+      "type": "unknown",
+      "version": "",
+      "author": "",
+      "description": "",
+      "downloads": 0,
+      "views": 236,
+      "upvotes": 7,
+      "saves": 3,
+      "comments": 2,
+      "created_at": "2026-01-10",
+      "updated_at": "2026-01-10",
+      "url": "https://openwebui.com/posts/debug_open_webui_plugins_in_your_browser_81bf7960"
+    }
+  ],
+  "user": {
+    "username": "Fu-Jie",
+    "name": "Fu-Jie",
+    "profile_url": "https://openwebui.com/u/Fu-Jie",
+    "profile_image": "https://community.s3.openwebui.com/uploads/users/b15d1348-4347-42b4-b815-e053342d6cb0/profile_d9510745-4bd4-4f8f-a997-4a21847d9300.webp",
+    "followers": 82,
+    "following": 2,
+    "total_points": 86,
+    "post_points": 74,
+    "comment_points": 12,
+    "contributions": 22
+  }
+}

docs/community-stats.md

@@ -0,0 +1,39 @@
+# 📊 OpenWebUI Community Stats Report
+
+> 📅 Updated: 2026-01-12 01:06
+
+## 📈 Overview
+
+| Metric | Value |
+|------|------|
+| 📝 Total Posts | 14 |
+| ⬇️ Total Downloads | 1161 |
+| 👁️ Total Views | 12713 |
+| 👍 Total Upvotes | 76 |
+| 💾 Total Saves | 73 |
+| 💬 Total Comments | 18 |
+
+## 📂 By Type
+
+- **action**: 11
+- **filter**: 2
+- **unknown**: 1
+
+## 📋 Posts List
+
+| Rank | Title | Type | Version | Downloads | Views | Upvotes | Saves | Updated |
+|:---:|------|:---:|:---:|:---:|:---:|:---:|:---:|:---:|
+| 1 | [Smart Mind Map](https://openwebui.com/posts/turn_any_text_into_beautiful_mind_maps_3094c59a) | action | 0.9.1 | 368 | 3352 | 11 | 22 | 2026-01-07 |
+| 2 | [Export to Excel](https://openwebui.com/posts/export_mulit_table_to_excel_244b8f9d) | action | 0.3.7 | 184 | 590 | 3 | 4 | 2026-01-07 |
+| 3 | [📊 Smart Infographic (AntV)](https://openwebui.com/posts/smart_infographic_ad6f0c7f) | action | 1.4.9 | 136 | 1482 | 8 | 9 | 2026-01-11 |
+| 4 | [Async Context Compression](https://openwebui.com/posts/async_context_compression_b1655bc8) | filter | 1.1.2 | 136 | 1510 | 6 | 10 | 2026-01-11 |
+| 5 | [Flash Card](https://openwebui.com/posts/flash_card_65a2ea8f) | action | 0.2.4 | 100 | 1833 | 8 | 6 | 2026-01-07 |
+| 6 | [Export to Word (Enhanced)](https://openwebui.com/posts/export_to_word_enhanced_formatting_fca6a315) | action | 0.4.3 | 96 | 880 | 6 | 9 | 2026-01-07 |
+| 7 | [📊 智能信息图 (AntV Infographic)](https://openwebui.com/posts/智能信息图_e04a48ff) | action | 1.4.9 | 35 | 511 | 3 | 0 | 2026-01-11 |
+| 8 | [导出为 Word (增强版)](https://openwebui.com/posts/导出为_word_支持公式流程图表格和代码块_8a6306c0) | action | 0.4.3 | 35 | 998 | 9 | 2 | 2026-01-07 |
+| 9 | [Deep Dive](https://openwebui.com/posts/deep_dive_c0b846e4) | action | 1.0.0 | 30 | 307 | 3 | 4 | 2026-01-08 |
+| 10 | [思维导图](https://openwebui.com/posts/智能生成交互式思维导图帮助用户可视化知识_8d4b097b) | action | 0.9.1 | 18 | 330 | 2 | 1 | 2026-01-07 |
+| 11 | [闪记卡 (Flash Card)](https://openwebui.com/posts/闪记卡生成插件_4a31eac3) | action | 0.2.4 | 12 | 361 | 4 | 1 | 2026-01-07 |
+| 12 | [异步上下文压缩](https://openwebui.com/posts/异步上下文压缩_5c0617cb) | filter | 1.1.2 | 8 | 211 | 4 | 1 | 2026-01-11 |
+| 13 | [精读](https://openwebui.com/posts/精读_99830b0f) | action | 1.0.0 | 3 | 112 | 2 | 1 | 2026-01-08 |
+| 14 | [ 🛠️ Debug Open WebUI Plugins in Your Browser](https://openwebui.com/posts/debug_open_webui_plugins_in_your_browser_81bf7960) | unknown |  | 0 | 236 | 7 | 3 | 2026-01-10 |

docs/community-stats.zh.md

@@ -0,0 +1,39 @@
+# 📊 OpenWebUI 社区统计报告
+
+> 📅 更新时间: 2026-01-12 01:06
+
+## 📈 总览
+
+| 指标 | 数值 |
+|------|------|
+| 📝 发布数量 | 14 |
+| ⬇️ 总下载量 | 1161 |
+| 👁️ 总浏览量 | 12713 |
+| 👍 总点赞数 | 76 |
+| 💾 总收藏数 | 73 |
+| 💬 总评论数 | 18 |
+
+## 📂 按类型分类
+
+- **action**: 11
+- **filter**: 2
+- **unknown**: 1
+
+## 📋 发布列表
+
+| 排名 | 标题 | 类型 | 版本 | 下载 | 浏览 | 点赞 | 收藏 | 更新日期 |
+|:---:|------|:---:|:---:|:---:|:---:|:---:|:---:|:---:|
+| 1 | [Smart Mind Map](https://openwebui.com/posts/turn_any_text_into_beautiful_mind_maps_3094c59a) | action | 0.9.1 | 368 | 3352 | 11 | 22 | 2026-01-07 |
+| 2 | [Export to Excel](https://openwebui.com/posts/export_mulit_table_to_excel_244b8f9d) | action | 0.3.7 | 184 | 590 | 3 | 4 | 2026-01-07 |
+| 3 | [📊 Smart Infographic (AntV)](https://openwebui.com/posts/smart_infographic_ad6f0c7f) | action | 1.4.9 | 136 | 1482 | 8 | 9 | 2026-01-11 |
+| 4 | [Async Context Compression](https://openwebui.com/posts/async_context_compression_b1655bc8) | filter | 1.1.2 | 136 | 1510 | 6 | 10 | 2026-01-11 |
+| 5 | [Flash Card](https://openwebui.com/posts/flash_card_65a2ea8f) | action | 0.2.4 | 100 | 1833 | 8 | 6 | 2026-01-07 |
+| 6 | [Export to Word (Enhanced)](https://openwebui.com/posts/export_to_word_enhanced_formatting_fca6a315) | action | 0.4.3 | 96 | 880 | 6 | 9 | 2026-01-07 |
+| 7 | [📊 智能信息图 (AntV Infographic)](https://openwebui.com/posts/智能信息图_e04a48ff) | action | 1.4.9 | 35 | 511 | 3 | 0 | 2026-01-11 |
+| 8 | [导出为 Word (增强版)](https://openwebui.com/posts/导出为_word_支持公式流程图表格和代码块_8a6306c0) | action | 0.4.3 | 35 | 998 | 9 | 2 | 2026-01-07 |
+| 9 | [Deep Dive](https://openwebui.com/posts/deep_dive_c0b846e4) | action | 1.0.0 | 30 | 307 | 3 | 4 | 2026-01-08 |
+| 10 | [思维导图](https://openwebui.com/posts/智能生成交互式思维导图帮助用户可视化知识_8d4b097b) | action | 0.9.1 | 18 | 330 | 2 | 1 | 2026-01-07 |
+| 11 | [闪记卡 (Flash Card)](https://openwebui.com/posts/闪记卡生成插件_4a31eac3) | action | 0.2.4 | 12 | 361 | 4 | 1 | 2026-01-07 |
+| 12 | [异步上下文压缩](https://openwebui.com/posts/异步上下文压缩_5c0617cb) | filter | 1.1.2 | 8 | 211 | 4 | 1 | 2026-01-11 |
+| 13 | [精读](https://openwebui.com/posts/精读_99830b0f) | action | 1.0.0 | 3 | 112 | 2 | 1 | 2026-01-08 |
+| 14 | [ 🛠️ Debug Open WebUI Plugins in Your Browser](https://openwebui.com/posts/debug_open_webui_plugins_in_your_browser_81bf7960) | unknown |  | 0 | 236 | 7 | 3 | 2026-01-10 |

docs/development/frontend-console-debugging.md

@@ -0,0 +1,150 @@
+# 🛠️ Debugging Python Plugins with Frontend Console
+
+When developing plugins for Open WebUI, debugging can be challenging. Standard `print()` statements or server-side logging might not always be accessible, especially in hosted environments or when you want to see the data flow in real-time alongside the UI interactions.
+
+This guide introduces a powerful technique: **Frontend Console Debugging**. By injecting JavaScript from your Python plugin, you can print structured logs directly to the browser's Developer Tools console (F12).
+
+## Why Frontend Debugging?
+
+*   **Real-time Feedback**: See logs immediately as actions happen in the browser.
+*   **Rich Objects**: Inspect complex JSON objects (like `body` or `messages`) interactively, rather than reading massive text dumps.
+*   **No Server Access Needed**: Debug issues even if you don't have SSH/Console access to the backend server.
+*   **Clean Output**: Group logs using `console.group()` to keep your console organized.
+
+## The Core Mechanism
+
+Open WebUI plugins (both Actions and Filters) support an event system. We can leverage the `__event_call__` (or sometimes `__event_emitter__`) to send a special event of type `execute`. This tells the frontend to run the provided JavaScript code.
+
+### The Helper Method
+
+To make this easy to use, we recommend adding a helper method `_emit_debug_log` to your plugin class.
+
+```python
+import json
+from typing import List
+
+async def _emit_debug_log(
+    self, 
+    __event_call__, 
+    title: str, 
+    data: dict
+):
+    """
+    Emit debug log to browser console via JS execution.
+    
+    Args:
+        __event_call__: The event callable passed to action/outlet.
+        title: A title for the log group.
+        data: A dictionary of data to log.
+    """
+    # 1. Check if debugging is enabled (recommended)
+    if not getattr(self.valves, "show_debug_log", True) or not __event_call__:
+        return
+
+    try:
+        # 2. Construct the JavaScript code
+        # We use an async IIFE (Immediately Invoked Function Expression) 
+        # to ensure a clean scope and support await if needed.
+        js_code = f"""
+            (async function() {{
+                console.group("🛠️ Plugin Debug: {title}");
+                console.log({json.dumps(data, ensure_ascii=False)});
+                console.groupEnd();
+            }})();
+        """
+
+        # 3. Send the execute event
+        await __event_call__(
+            {
+                "type": "execute",
+                "data": {"code": js_code},
+            }
+        )
+    except Exception as e:
+        print(f"Error emitting debug log: {e}")
+```
+
+## Implementation Steps
+
+### 1. Add a Valve for Control
+
+It's best practice to make debugging optional so it doesn't clutter the console for normal users.
+
+```python
+from pydantic import BaseModel, Field
+
+class Filter:
+    class Valves(BaseModel):
+        show_debug_log: bool = Field(
+            default=False, 
+            description="Print debug logs to browser console (F12)"
+        )
+    
+    def __init__(self):
+        self.valves = self.Valves()
+```
+
+### 2. Inject `__event_call__`
+
+Ensure your `action` (for Actions) or `outlet` (for Filters) method accepts `__event_call__`.
+
+**For Filters (`outlet`):**
+
+```python
+async def outlet(
+    self,
+    body: dict,
+    __user__: Optional[dict] = None,
+    __event_call__=None,  # <--- Add this
+    __metadata__: Optional[dict] = None,
+) -> dict:
+```
+
+**For Actions (`action`):**
+
+```python
+async def action(
+    self,
+    body: dict,
+    __user__=None,
+    __event_call__=None, # <--- Add this
+    __request__=None,
+):
+```
+
+### 3. Call the Helper
+
+Now you can log anything, anywhere in your logic!
+
+```python
+# Inside your logic...
+new_content = self.process_content(content)
+
+# Log the before and after
+await self._emit_debug_log(
+    __event_call__,
+    "Content Normalization",
+    {
+        "original": content,
+        "processed": new_content,
+        "changes": diff_list
+    }
+)
+```
+
+## Best Practices
+
+1.  **Use `json.dumps`**: Always serialize your Python dictionaries to JSON strings before embedding them in the f-string. This handles escaping quotes and special characters correctly.
+2.  **Async IIFE**: Wrapping your JS in `(async function() { ... })();` is safer than raw code. It prevents variable collisions with other scripts and allows using `await` inside your debug script if you ever need to check DOM elements.
+3.  **Check for None**: Always check if `__event_call__` is not None before using it, as it might not be available in all contexts (e.g., when running tests or in older Open WebUI versions).
+
+## Example Output
+
+When enabled, your browser console will show:
+
+```text
+> 🛠️ Plugin Debug: Content Normalization
+  > {original: "...", processed: "...", changes: [...]}
+```
+
+You can expand the object to inspect every detail of your data. Happy debugging!

docs/development/mermaid-syntax-standards.md

@@ -0,0 +1,64 @@
+# Mermaid Syntax Standards & Best Practices
+
+This document summarizes the official syntax standards for Mermaid flowcharts, focusing on node labels, quoting rules, and special character handling. It serves as a reference for the `markdown_normalizer` plugin logic.
+
+## 1. Node Shapes & Syntax
+
+Mermaid supports various node shapes defined by specific wrapping characters.
+
+| Shape | Syntax | Example |
+| :--- | :--- | :--- |
+| **Rectangle** (Default) | `id[Label]` | `A[Start]` |
+| **Rounded** | `id(Label)` | `B(Process)` |
+| **Stadium** (Pill) | `id([Label])` | `C([End])` |
+| **Subroutine** | `id[[Label]]` | `D[[Subroutine]]` |
+| **Cylinder** (Database) | `id[(Label)]` | `E[(Database)]` |
+| **Circle** | `id((Label))` | `F((Point))` |
+| **Double Circle** | `id(((Label)))` | `G(((Endpoint)))` |
+| **Asymmetric** | `id>Label]` | `H>Flag]` |
+| **Rhombus** (Decision) | `id{Label}` | `I{Decision}` |
+| **Hexagon** | `id{{Label}}` | `J{{Prepare}}` |
+| **Parallelogram** | `id[/Label/]` | `K[/Input/]` |
+| **Parallelogram Alt** | `id[\Label\]` | `L[\Output\]` |
+| **Trapezoid** | `id[/Label\]` | `M[/Trap/]` |
+| **Trapezoid Alt** | `id[\Label/]` | `N[\TrapAlt/]` |
+
+## 2. Quoting Rules (Critical)
+
+### Why Quote?
+Quoting node labels is **highly recommended** and sometimes **mandatory** to prevent syntax errors.
+
+### Mandatory Quoting Scenarios
+You **MUST** enclose labels in double quotes `"` if they contain:
+1.  **Special Characters**: `()`, `[]`, `{}`, `;`, `"`, etc.
+2.  **Keywords**: Words like `end`, `subgraph`, etc., if used in specific contexts.
+3.  **Unicode/Emoji**: While often supported without quotes, quoting ensures consistent rendering across different environments.
+4.  **Markdown**: If you want to use Markdown formatting (bold, italic) inside a label.
+
+### Best Practice: Always Quote
+To ensure robustness, especially when processing LLM-generated content which may contain unpredictable characters, **always enclosing labels in double quotes is the safest strategy**.
+
+**Examples:**
+*   ❌ Risky: `id(Start: 15:00)` (Colon might be interpreted as style separator)
+*   ✅ Safe: `id("Start: 15:00")`
+*   ❌ Broken: `id(Func(x))` (Nested parentheses break parsing)
+*   ✅ Safe: `id("Func(x)")`
+
+## 3. Escape Characters
+
+Inside a quoted string:
+*   Double quotes `"` must be escaped as `\"`.
+*   HTML entities (e.g., `#35;` for `#`) can be used.
+
+## 4. Plugin Logic Verification
+
+The `markdown_normalizer` plugin implements the following logic:
+
+1.  **Detection**: Identifies Mermaid node definitions using a comprehensive regex covering all shapes above.
+2.  **Normalization**:
+    *   Checks if the label is already quoted.
+    *   If **NOT quoted**, it wraps the label in double quotes `""`.
+    *   Escapes any existing double quotes inside the label (`"` -> `\"`).
+3.  **Shape Preservation**: The regex captures the specific opening and closing delimiters (e.g., `((` and `))`) to ensure the node shape is strictly preserved during normalization.
+
+**Conclusion**: The plugin's behavior of automatically adding quotes to unquoted labels is **fully aligned with Mermaid's official best practices** for robustness and error prevention.

docs/development/plugin-guide.md

@@ -235,6 +235,125 @@ llm_response = await generate_chat_completion(
 )
 ```
 
+### 4.4 JS Render to Markdown (Data URL Embedding)
+
+For scenarios requiring complex frontend rendering (e.g., AntV charts, Mermaid diagrams) but wanting **persistent pure Markdown output**, use the Data URL embedding pattern:
+
+#### Workflow
+
+```
+┌──────────────────────────────────────────────────────────────┐
+│  1. Python Action                                             │
+│     ├── Analyze message content                               │
+│     ├── Call LLM to generate structured data (optional)       │
+│     └── Send JS code to frontend via __event_call__           │
+├──────────────────────────────────────────────────────────────┤
+│  2. Browser JS (via __event_call__)                           │
+│     ├── Dynamically load visualization library                │
+│     ├── Render SVG/Canvas offscreen                           │
+│     ├── Export to Base64 Data URL via toDataURL()             │
+│     └── Update message content via REST API                   │
+├──────────────────────────────────────────────────────────────┤
+│  3. Markdown Rendering                                        │
+│     └── Display ![description](data:image/svg+xml;base64,...) │
+└──────────────────────────────────────────────────────────────┘
+```
+
+#### Python Side (Send JS for Execution)
+
+```python
+async def action(self, body, __event_call__, __metadata__, ...):
+    chat_id = self._extract_chat_id(body, __metadata__)
+    message_id = self._extract_message_id(body, __metadata__)
+    
+    # Generate JS code
+    js_code = self._generate_js_code(
+        chat_id=chat_id,
+        message_id=message_id,
+        data=processed_data,
+    )
+    
+    # Execute JS
+    if __event_call__:
+        await __event_call__({
+            "type": "execute",
+            "data": {"code": js_code}
+        })
+```
+
+#### JavaScript Side (Render and Write-back)
+
+```javascript
+(async function() {
+    // 1. Load visualization library
+    if (typeof VisualizationLib === 'undefined') {
+        await new Promise((resolve, reject) => {
+            const script = document.createElement('script');
+            script.src = 'https://cdn.example.com/lib.min.js';
+            script.onload = resolve;
+            script.onerror = reject;
+            document.head.appendChild(script);
+        });
+    }
+    
+    // 2. Create offscreen container
+    const container = document.createElement('div');
+    container.style.cssText = 'position:absolute;left:-9999px;';
+    document.body.appendChild(container);
+    
+    // 3. Render visualization
+    const instance = new VisualizationLib({ container });
+    instance.render(data);
+    
+    // 4. Export to Data URL
+    const dataUrl = await instance.toDataURL({ type: 'svg', embedResources: true });
+    
+    // 5. Cleanup
+    instance.destroy();
+    document.body.removeChild(container);
+    
+    // 6. Generate Markdown image
+    const markdownImage = `![Chart](${dataUrl})`;
+    
+    // 7. Update message via API
+    const token = localStorage.getItem("token");
+    await fetch(`/api/v1/chats/${chatId}/messages/${messageId}/event`, {
+        method: "POST",
+        headers: {
+            "Content-Type": "application/json",
+            "Authorization": `Bearer ${token}`
+        },
+        body: JSON.stringify({
+            type: "chat:message",
+            data: { content: originalContent + "\n\n" + markdownImage }
+        })
+    });
+})();
+```
+
+#### Benefits
+
+- **Pure Markdown Output**: Standard Markdown image syntax, no HTML code blocks
+- **Self-Contained**: Images embedded as Base64 Data URL, no external dependencies
+- **Persistent**: Via API write-back, images remain after page reload
+- **Cross-Platform**: Works on any client supporting Markdown images
+
+#### HTML Injection vs JS Render to Markdown
+
+| Feature | HTML Injection | JS Render + Markdown |
+|---------|----------------|----------------------|
+| Output Format | HTML code block | Markdown image |
+| Interactivity | ✅ Buttons, animations | ❌ Static image |
+| External Deps | Requires JS libraries | None (self-contained) |
+| Persistence | Depends on browser | ✅ Permanent |
+| File Export | Needs special handling | ✅ Direct export |
+| Use Case | Interactive content | Infographics, chart snapshots |
+
+#### Reference Implementations
+
+- `plugins/actions/js-render-poc/infographic_markdown.py` - AntV Infographic + Data URL
+- `plugins/actions/js-render-poc/js_render_poc.py` - Basic proof of concept
+
 ---
 
 ## 5. Best Practices & Design Principles

docs/development/plugin-guide.zh.md

@@ -199,7 +199,124 @@ async def background_job(self, chat_id):
     pass
 ```
 
----
+### 4.3 JS 渲染并嵌入 Markdown (Data URL 嵌入)
+
+对于需要复杂前端渲染（如 AntV 图表、Mermaid 图表）但希望结果**持久化为纯 Markdown 格式**的场景，推荐使用 Data URL 嵌入模式：
+
+#### 工作流程
+
+```
+┌──────────────────────────────────────────────────────────────┐
+│  1. Python Action                                             │
+│     ├── 分析消息内容                                           │
+│     ├── 调用 LLM 生成结构化数据（可选）                         │
+│     └── 通过 __event_call__ 发送 JS 代码到前端                 │
+├──────────────────────────────────────────────────────────────┤
+│  2. Browser JS (通过 __event_call__)                          │
+│     ├── 动态加载可视化库                                       │
+│     ├── 离屏渲染 SVG/Canvas                                    │
+│     ├── 使用 toDataURL() 导出 Base64 Data URL                  │
+│     └── 通过 REST API 更新消息内容                             │
+├──────────────────────────────────────────────────────────────┤
+│  3. Markdown 渲染                                             │
+│     └── 显示 ![描述](data:image/svg+xml;base64,...)           │
+└──────────────────────────────────────────────────────────────┘
+```
+
+#### Python 端（发送 JS 执行）
+
+```python
+async def action(self, body, __event_call__, __metadata__, ...):
+    chat_id = self._extract_chat_id(body, __metadata__)
+    message_id = self._extract_message_id(body, __metadata__)
+    
+    # 生成 JS 代码
+    js_code = self._generate_js_code(
+        chat_id=chat_id,
+        message_id=message_id,
+        data=processed_data,
+    )
+    
+    # 执行 JS
+    if __event_call__:
+        await __event_call__({
+            "type": "execute",
+            "data": {"code": js_code}
+        })
+```
+
+#### JavaScript 端（渲染并回写）
+
+```javascript
+(async function() {
+    // 1. 加载可视化库
+    if (typeof VisualizationLib === 'undefined') {
+        await new Promise((resolve, reject) => {
+            const script = document.createElement('script');
+            script.src = 'https://cdn.example.com/lib.min.js';
+            script.onload = resolve;
+            script.onerror = reject;
+            document.head.appendChild(script);
+        });
+    }
+    
+    // 2. 创建离屏容器
+    const container = document.createElement('div');
+    container.style.cssText = 'position:absolute;left:-9999px;';
+    document.body.appendChild(container);
+    
+    // 3. 渲染可视化
+    const instance = new VisualizationLib({ container });
+    instance.render(data);
+    
+    // 4. 导出为 Data URL
+    const dataUrl = await instance.toDataURL({ type: 'svg', embedResources: true });
+    
+    // 5. 清理
+    instance.destroy();
+    document.body.removeChild(container);
+    
+    // 6. 生成 Markdown 图片
+    const markdownImage = `![图表](${dataUrl})`;
+    
+    // 7. 通过 API 更新消息
+    const token = localStorage.getItem("token");
+    await fetch(`/api/v1/chats/${chatId}/messages/${messageId}/event`, {
+        method: "POST",
+        headers: {
+            "Content-Type": "application/json",
+            "Authorization": `Bearer ${token}`
+        },
+        body: JSON.stringify({
+            type: "chat:message",
+            data: { content: originalContent + "\n\n" + markdownImage }
+        })
+    });
+})();
+```
+
+#### 优势
+
+- **纯 Markdown 输出**：结果是标准的 Markdown 图片语法，无需 HTML 代码块
+- **自包含**：图片以 Base64 Data URL 嵌入，无外部依赖
+- **持久化**：通过 API 回写，消息重新加载后图片仍然存在
+- **跨平台**：任何支持 Markdown 图片的客户端都能显示
+
+#### HTML 注入 vs JS 渲染嵌入 Markdown
+
+| 特性 | HTML 注入 | JS 渲染 + Markdown 图片 |
+|------|----------|------------------------|
+| 输出格式 | HTML 代码块 | Markdown 图片 |
+| 交互性 | ✅ 支持按钮、动画 | ❌ 静态图片 |
+| 外部依赖 | 需要加载 JS 库 | 无（图片自包含） |
+| 持久化 | 依赖浏览器渲染 | ✅ 永久可见 |
+| 文件导出 | 需特殊处理 | ✅ 直接导出 |
+| 适用场景 | 交互式内容 | 信息图、图表快照 |
+
+#### 参考实现
+
+- `plugins/actions/js-render-poc/infographic_markdown.py` - AntV 信息图 + Data URL
+- `plugins/actions/js-render-poc/js_render_poc.py` - 基础概念验证
 
 ## 5. 最佳实践与设计原则
 

docs/examples/action_plugin_smart_mind_map_example_cn.md

@@ -21,7 +21,7 @@
 
 Open WebUI 通过文件顶部的特定格式注释来识别和展示插件信息。
 
-**代码示例 (`思维导图.py`):**
+**代码示例 (`smart_mind_map_cn.py`):**
 
 ```python
 """
@@ -45,7 +45,7 @@ description: 智能分析文本内容,生成交互式思维导图,帮助用户
 
 通过在 `Action` 类内部定义一个 `Valves` Pydantic 模型，可以为插件创建可在 Web UI 中配置的参数。
 
-**代码示例 (`思维导图.py`):**
+**代码示例 (`smart_mind_map_cn.py`):**
 
 ```python
 class Action:
@@ -83,7 +83,7 @@ class Action:
 
 `action` 方法是插件的执行入口，它是一个异步函数，接收 Open WebUI 传入的上下文信息。
 
-**代码示例 (`思维导图.py`):**
+**代码示例 (`smart_mind_map_cn.py`):**
 
 ```python
     async def action(

docs/index.md

@@ -104,10 +104,16 @@ hide:
 
 ### Using Plugins
 
-1. Browse the [Plugin Center](plugins/index.md) and download the plugin file (`.py`)
-2. Open OpenWebUI **Admin Panel** → **Settings** → **Plugins**
-3. Click the upload button and select the `.py` file
-4. Refresh the page and enable the plugin in your chat settings
+1. **Install from OpenWebUI Community (Recommended)**:
+   - Visit my profile: [Fu-Jie's Profile](https://openwebui.com/u/Fu-Jie)
+   - Browse the plugins and select the one you like.
+   - Click "Get" to import it directly into your OpenWebUI instance.
+
+2. **Manual Installation**:
+   - Browse the [Plugin Center](plugins/index.md) and download the plugin file (`.py`)
+   - Open OpenWebUI **Admin Panel** → **Settings** → **Plugins**
+   - Click the upload button and select the `.py` file
+   - Refresh the page and enable the plugin in your chat settings
 
 ---
 

docs/index.zh.md

@@ -104,10 +104,16 @@ hide:
 
 ### 使用插件
 
-1. 浏览[插件中心](plugins/index.md)并下载插件文件（`.py`）
-2. 打开 OpenWebUI **管理面板** → **设置** → **插件**
-3. 点击上传按钮并选择 `.py` 文件
-4. 刷新页面并在聊天设置中启用插件
+1. **从 OpenWebUI 社区安装 (推荐)**:
+   - 访问我的主页: [Fu-Jie's Profile](https://openwebui.com/u/Fu-Jie)
+   - 浏览插件列表，选择你喜欢的插件。
+   - 点击 "Get" 按钮，将其直接导入到你的 OpenWebUI 实例中。
+
+2. **手动安装**:
+   - 浏览[插件中心](plugins/index.md)并下载插件文件（`.py`）
+   - 打开 OpenWebUI **管理面板** → **设置** → **插件**
+   - 点击上传按钮并选择 `.py` 文件
+   - 刷新页面并在聊天设置中启用插件
 
 ---
 

docs/js-visualization-guide.md

@@ -0,0 +1,290 @@
+# 使用 JavaScript 生成可视化内容的技术方案
+
+## 概述
+
+本文档描述了在 OpenWebUI Action 插件中使用浏览器端 JavaScript 代码生成可视化内容（如思维导图、信息图等）并将结果保存到消息中的技术方案。
+
+## 核心架构
+
+```mermaid
+sequenceDiagram
+    participant Plugin as Python 插件
+    participant EventCall as __event_call__
+    participant Browser as 浏览器 (JS)
+    participant API as OpenWebUI API
+    participant DB as 数据库
+
+    Plugin->>EventCall: 1. 发送 execute 事件 (含 JS 代码)
+    EventCall->>Browser: 2. 执行 JS 代码
+    Browser->>Browser: 3. 加载可视化库 (D3/Markmap/AntV)
+    Browser->>Browser: 4. 渲染可视化内容
+    Browser->>Browser: 5. 转换为 Base64 Data URI
+    Browser->>API: 6. GET 获取当前消息内容
+    API-->>Browser: 7. 返回消息数据
+    Browser->>API: 8. POST 追加 Markdown 图片到消息
+    API->>DB: 9. 保存更新后的消息
+```
+
+## 关键步骤
+
+### 1. Python 端通过 `__event_call__` 执行 JS
+
+Python 插件**不直接修改 `body["messages"]`**，而是通过 `__event_call__` 发送 JS 代码让浏览器执行：
+
+```python
+async def action(
+    self,
+    body: dict,
+    __user__: dict = None,
+    __event_emitter__=None,
+    __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+    __metadata__: Optional[dict] = None,
+    __request__: Request = None,
+) -> dict:
+    # 从 body 获取 chat_id 和 message_id
+    chat_id = body.get("chat_id", "")
+    message_id = body.get("id", "")  # 注意：body["id"] 是 message_id
+    
+    # 通过 __event_call__ 执行 JS 代码
+    if __event_call__:
+        await __event_call__({
+            "type": "execute",
+            "data": {
+                "code": f"""
+(async function() {{
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    // ... JS 渲染和 API 更新逻辑 ...
+}})();
+                """
+            },
+        })
+    
+    # 不修改 body，直接返回
+    return body
+```
+
+### 2. JavaScript 加载可视化库
+
+在浏览器端动态加载所需的 JS 库：
+
+```javascript
+// 加载 D3.js
+if (!window.d3) {
+    await new Promise((resolve, reject) => {
+        const script = document.createElement('script');
+        script.src = 'https://cdn.jsdelivr.net/npm/d3@7';
+        script.onload = resolve;
+        script.onerror = reject;
+        document.head.appendChild(script);
+    });
+}
+
+// 加载 Markmap (思维导图)
+if (!window.markmap) {
+    await loadScript('https://cdn.jsdelivr.net/npm/markmap-lib@0.17');
+    await loadScript('https://cdn.jsdelivr.net/npm/markmap-view@0.17');
+}
+```
+
+### 3. 渲染并转换为 Data URI
+
+```javascript
+// 创建 SVG 元素
+const svg = document.createElementNS('http://www.w3.org/2000/svg', 'svg');
+svg.setAttribute('width', '800');
+svg.setAttribute('height', '600');
+svg.setAttribute('xmlns', 'http://www.w3.org/2000/svg');
+
+// ... 执行渲染逻辑 (添加图形元素) ...
+
+// 转换为 Base64 Data URI
+const svgData = new XMLSerializer().serializeToString(svg);
+const base64 = btoa(unescape(encodeURIComponent(svgData)));
+const dataUri = 'data:image/svg+xml;base64,' + base64;
+```
+
+### 4. 获取当前消息内容
+
+由于 Python 端不传递原始内容，JS 需要通过 API 获取：
+
+```javascript
+const token = localStorage.getItem('token');
+
+// 获取当前聊天数据
+const getResponse = await fetch(`/api/v1/chats/${chatId}`, {
+    method: 'GET',
+    headers: {
+        'Authorization': `Bearer ${token}`
+    }
+});
+
+const chatData = await getResponse.json();
+
+// 查找目标消息
+let originalContent = '';
+if (chatData.chat && chatData.chat.messages) {
+    const targetMsg = chatData.chat.messages.find(m => m.id === messageId);
+    if (targetMsg && targetMsg.content) {
+        originalContent = targetMsg.content;
+    }
+}
+```
+
+### 5. 调用 API 更新消息
+
+```javascript
+// 构造新内容：原始内容 + Markdown 图片
+const markdownImage = `![可视化图片](${dataUri})`;
+const newContent = originalContent + '\n\n' + markdownImage;
+
+// 调用 API 更新消息
+const response = await fetch(`/api/v1/chats/${chatId}/messages/${messageId}/event`, {
+    method: 'POST',
+    headers: {
+        'Content-Type': 'application/json',
+        'Authorization': `Bearer ${token}`
+    },
+    body: JSON.stringify({
+        type: 'chat:message',
+        data: { content: newContent }
+    })
+});
+
+if (response.ok) {
+    console.log('消息更新成功！');
+}
+```
+
+## 完整示例
+
+参考 [js_render_poc.py](https://github.com/Fu-Jie/awesome-openwebui/blob/main/plugins/actions/js-render-poc/js_render_poc.py) 获取完整的 PoC 实现。
+
+## 事件类型
+
+| 类型 | 用途 |
+|------|------|
+| `chat:message:delta` | 增量更新（追加文本） |
+| `chat:message` | 完全替换消息内容 |
+
+```javascript
+// 增量更新
+{ type: "chat:message:delta", data: { content: "追加的内容" } }
+
+// 完全替换
+{ type: "chat:message", data: { content: "完整的新内容" } }
+```
+
+## 关键数据来源
+
+| 数据 | 来源 | 说明 |
+|------|------|------|
+| `chat_id` | `body["chat_id"]` | 聊天会话 ID |
+| `message_id` | `body["id"]` | ⚠️ 注意：是 `body["id"]`，不是 `body["message_id"]` |
+| `token` | `localStorage.getItem('token')` | 用户认证 Token |
+| `originalContent` | 通过 API `GET /api/v1/chats/{chatId}` 获取 | 当前消息内容 |
+
+## Python 端 API
+
+| 参数 | 类型 | 说明 |
+|------|------|------|
+| `__event_emitter__` | Callable | 发送状态/通知事件 |
+| `__event_call__` | Callable | 执行 JS 代码（用于可视化渲染） |
+| `__metadata__` | dict | 元数据（可能为 None） |
+| `body` | dict | 请求体，包含 messages、chat_id、id 等 |
+
+### body 结构示例
+
+```json
+{
+  "model": "gemini-3-flash-preview",
+  "messages": [...],
+  "chat_id": "ac2633a3-5731-4944-98e3-bf9b3f0ef0ab",
+  "id": "2e0bb7d4-dfc0-43d7-b028-fd9e06c6fdc8",
+  "session_id": "bX30sHI8r4_CKxCdAAAL"
+}
+```
+
+### 常用事件
+
+```python
+# 发送状态更新
+await __event_emitter__({
+    "type": "status",
+    "data": {"description": "正在渲染...", "done": False}
+})
+
+# 执行 JS 代码
+await __event_call__({
+    "type": "execute",
+    "data": {"code": "console.log('Hello from Python!')"}
+})
+
+# 发送通知
+await __event_emitter__({
+    "type": "notification",
+    "data": {"type": "success", "content": "渲染完成！"}
+})
+```
+
+## 适用场景
+
+- **思维导图** (Markmap)
+- **信息图** (AntV Infographic)
+- **流程图** (Mermaid)
+- **数据图表** (ECharts, Chart.js)
+- **任何需要 JS 渲染的可视化内容**
+
+## 注意事项
+
+### 1. 竞态条件问题
+
+⚠️ **多次快速点击会导致内容覆盖问题**
+
+由于 API 调用是异步的，如果用户快速多次触发 Action：
+- 第一次点击：获取原始内容 A → 渲染 → 更新为 A+图片1
+- 第二次点击：可能获取到旧内容 A（第一次还没保存完）→ 更新为 A+图片2
+
+结果：图片1 被覆盖丢失！
+
+**解决方案**：
+- 添加防抖（debounce）机制
+- 使用锁/标志位防止重复执行
+- 或使用 `chat:message:delta` 增量更新
+
+### 2. 不要直接修改 `body["messages"]`
+
+消息更新应由 JS 通过 API 完成，确保获取最新内容。
+
+### 3. f-string 限制
+
+Python f-string 内不能直接使用反斜杠，需要将转义字符串预先处理：
+
+```python
+# 转义 JSON 中的特殊字符
+body_json = json.dumps(data, ensure_ascii=False)
+escaped = body_json.replace("\\", "\\\\").replace("`", "\\`").replace("${", "\\${")
+```
+
+### 4. Data URI 大小限制
+
+Base64 编码会增加约 33% 的体积，复杂图片可能导致消息过大。
+
+### 5. 跨域问题
+
+确保 CDN 资源支持 CORS。
+
+### 6. API 权限
+
+确保用户 token 有权限访问和更新目标消息。
+
+## 与传统方式对比
+
+| 特性 | 传统方式 (修改 body) | 新方式 (__event_call__) |
+|------|---------------------|------------------------|
+| 消息更新 | Python 直接修改 | JS 通过 API 更新 |
+| 原始内容 | Python 传递给 JS | JS 通过 API 获取 |
+| 灵活性 | 低 | 高 |
+| 实时性 | 一次性 | 可多次更新 |
+| 复杂度 | 简单 | 中等 |
+| 竞态风险 | 低 | ⚠️ 需要处理 |

docs/plugins/actions/deep-dive.md

@@ -0,0 +1,111 @@
+# Deep Dive
+
+<span class="category-badge action">Action</span>
+<span class="version-badge">v1.0.0</span>
+
+A comprehensive thinking lens that dives deep into any content - from context to logic, insights, and action paths.
+
+---
+
+## Overview
+
+The Deep Dive plugin transforms how you understand complex content by guiding you through a structured thinking process. Rather than just summarizing, it deconstructs content across four phases:
+
+- **🔍 The Context (What?)**: Panoramic view of the situation and background
+- **🧠 The Logic (Why?)**: Deconstruction of reasoning and mental models  
+- **💎 The Insight (So What?)**: Non-obvious value and hidden implications
+- **🚀 The Path (Now What?)**: Specific, prioritized strategic actions
+
+## Features
+
+- :material-brain: **Thinking Chain**: Complete structured analysis process
+- :material-eye: **Deep Understanding**: Reveals hidden assumptions and blind spots
+- :material-lightbulb-on: **Insight Extraction**: Finds the "Aha!" moments
+- :material-rocket-launch: **Action Oriented**: Translates understanding into actionable steps
+- :material-theme-light-dark: **Theme Adaptive**: Auto-adapts to OpenWebUI light/dark theme
+- :material-translate: **Multi-language**: Outputs in user's preferred language
+
+---
+
+## Installation
+
+1. Download the plugin file: [`deep_dive.py`](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/deep-dive)
+2. Upload to OpenWebUI: **Admin Panel** → **Settings** → **Functions**
+3. Enable the plugin
+
+---
+
+## Usage
+
+1. Provide any long text, article, or meeting notes in the chat
+2. Click the **Deep Dive** button in the message action bar
+3. Follow the visual timeline from Context → Logic → Insight → Path
+
+---
+
+## Configuration
+
+| Option | Type | Default | Description |
+|--------|------|---------|-------------|
+| `SHOW_STATUS` | boolean | `true` | Show status updates during processing |
+| `MODEL_ID` | string | `""` | LLM model for analysis (empty = current model) |
+| `MIN_TEXT_LENGTH` | integer | `200` | Minimum text length for analysis |
+| `CLEAR_PREVIOUS_HTML` | boolean | `true` | Clear previous plugin results |
+| `MESSAGE_COUNT` | integer | `1` | Number of recent messages to analyze |
+
+---
+
+## Theme Support
+
+Deep Dive automatically adapts to OpenWebUI's light/dark theme:
+
+- Detects theme from parent document `<meta name="theme-color">` tag
+- Falls back to `html/body` class or `data-theme` attribute
+- Uses system preference `prefers-color-scheme: dark` as last resort
+
+!!! tip "For Best Results"
+    Enable **iframe Sandbox Allow Same Origin** in OpenWebUI:
+    **Settings** → **Interface** → **Artifacts** → Check **iframe Sandbox Allow Same Origin**
+
+---
+
+## Example Output
+
+The plugin generates a beautiful structured timeline:
+
+```
+┌─────────────────────────────────────┐
+│  🌊 Deep Dive Analysis              │
+│  👤 User  📅 Date  📊 Word count    │
+├─────────────────────────────────────┤
+│  🔍 Phase 01: The Context           │
+│  [High-level panoramic view]        │
+│                                     │
+│  🧠 Phase 02: The Logic             │
+│  • Reasoning structure...           │
+│  • Hidden assumptions...            │
+│                                     │
+│  💎 Phase 03: The Insight           │
+│  • Non-obvious value...             │
+│  • Blind spots revealed...          │
+│                                     │
+│  🚀 Phase 04: The Path              │
+│  ▸ Priority Action 1...             │
+│  ▸ Priority Action 2...             │
+└─────────────────────────────────────┘
+```
+
+---
+
+## Requirements
+
+!!! note "Prerequisites"
+    - OpenWebUI v0.3.0 or later
+    - Uses the active LLM model for analysis
+    - Requires `markdown` Python package
+
+---
+
+## Source Code
+
+[:fontawesome-brands-github: View on GitHub](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/deep-dive){ .md-button }

docs/plugins/actions/deep-dive.zh.md

@@ -0,0 +1,111 @@
+# 精读 (Deep Dive)
+
+<span class="category-badge action">Action</span>
+<span class="version-badge">v1.0.0</span>
+
+全方位的思维透镜 —— 从背景全景到逻辑脉络，从深度洞察到行动路径。
+
+---
+
+## 概述
+
+精读插件改变了您理解复杂内容的方式，通过结构化的思维过程引导您进行深度分析。它不仅仅是摘要，而是从四个阶段解构内容：
+
+- **🔍 全景 (The Context)**: 情境与背景的高层级全景视图
+- **🧠 脉络 (The Logic)**: 解构底层推理逻辑与思维模型
+- **💎 洞察 (The Insight)**: 提取非显性价值与隐藏含义
+- **🚀 路径 (The Path)**: 具体的、按优先级排列的战略行动
+
+## 功能特性
+
+- :material-brain: **思维链**: 完整的结构化分析过程
+- :material-eye: **深度理解**: 揭示隐藏的假设和思维盲点
+- :material-lightbulb-on: **洞察提取**: 发现"原来如此"的时刻
+- :material-rocket-launch: **行动导向**: 将深度理解转化为可执行步骤
+- :material-theme-light-dark: **主题自适应**: 自动适配 OpenWebUI 深色/浅色主题
+- :material-translate: **多语言**: 以用户偏好语言输出
+
+---
+
+## 安装
+
+1. 下载插件文件: [`deep_dive_cn.py`](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/deep-dive)
+2. 上传到 OpenWebUI: **管理面板** → **设置** → **Functions**
+3. 启用插件
+
+---
+
+## 使用方法
+
+1. 在聊天中提供任何长文本、文章或会议记录
+2. 点击消息操作栏中的 **精读** 按钮
+3. 沿着视觉时间轴从"全景"探索到"路径"
+
+---
+
+## 配置参数
+
+| 选项 | 类型 | 默认值 | 描述 |
+|------|------|--------|------|
+| `SHOW_STATUS` | boolean | `true` | 处理过程中是否显示状态更新 |
+| `MODEL_ID` | string | `""` | 用于分析的 LLM 模型（空 = 当前模型） |
+| `MIN_TEXT_LENGTH` | integer | `200` | 分析所需的最小文本长度 |
+| `CLEAR_PREVIOUS_HTML` | boolean | `true` | 是否清除之前的插件结果 |
+| `MESSAGE_COUNT` | integer | `1` | 要分析的最近消息数量 |
+
+---
+
+## 主题支持
+
+精读插件自动适配 OpenWebUI 的深色/浅色主题：
+
+- 从父文档 `<meta name="theme-color">` 标签检测主题
+- 回退到 `html/body` 的 class 或 `data-theme` 属性
+- 最后使用系统偏好 `prefers-color-scheme: dark`
+
+!!! tip "最佳效果"
+    请在 OpenWebUI 中启用 **iframe Sandbox Allow Same Origin**：
+    **设置** → **界面** → **Artifacts** → 勾选 **iframe Sandbox Allow Same Origin**
+
+---
+
+## 输出示例
+
+插件生成精美的结构化时间轴：
+
+```
+┌─────────────────────────────────────┐
+│  📖 精读分析报告                     │
+│  👤 用户  📅 日期  📊 字数           │
+├─────────────────────────────────────┤
+│  🔍 阶段 01: 全景 (The Context)      │
+│  [高层级全景视图内容]                 │
+│                                     │
+│  🧠 阶段 02: 脉络 (The Logic)        │
+│  • 推理结构分析...                   │
+│  • 隐藏假设识别...                   │
+│                                     │
+│  💎 阶段 03: 洞察 (The Insight)      │
+│  • 非显性价值提取...                 │
+│  • 思维盲点揭示...                   │
+│                                     │
+│  🚀 阶段 04: 路径 (The Path)         │
+│  ▸ 优先级行动 1...                   │
+│  ▸ 优先级行动 2...                   │
+└─────────────────────────────────────┘
+```
+
+---
+
+## 系统要求
+
+!!! note "前提条件"
+    - OpenWebUI v0.3.0 或更高版本
+    - 使用当前活跃的 LLM 模型进行分析
+    - 需要 `markdown` Python 包
+
+---
+
+## 源代码
+
+[:fontawesome-brands-github: 在 GitHub 上查看](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/deep-dive){ .md-button }

docs/plugins/actions/export-to-excel.md

@@ -1,12 +1,25 @@
 # Export to Excel
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v0.3.3</span>
+<span class="version-badge">v0.3.7</span>
 
 Export chat conversations to Excel spreadsheet format for analysis, archiving, and sharing.
 
+
+### What's New in v0.3.6
+- **OpenWebUI-Style Theme**: Modern dark header with light gray zebra striping for better readability.
+- **Zebra Striping**: Alternating row colors for improved visual scanning.
+- **Smart Data Type Conversion**: Automatically converts columns to numeric or datetime types.
+- **Full Cell Bold/Italic**: Supports Markdown bold/italic formatting in Excel.
+- **Partial Markdown Cleanup**: Removes partial Markdown symbols for cleaner output.
+- **Export Scope**: Choose between "Last Message" or "All Messages".
+- **Smart Sheet Naming**: Names sheets based on Markdown headers or message index.
+- **Smart Filename Generation**: Generates filenames based on Chat Title, AI Summary, or Markdown Headers.
+- **AI Title Generation**: Supports using a specific model (`MODEL_ID`) for title generation with progress notifications.
+
 ---
 
+
 ## Overview
 
 The Export to Excel plugin allows you to download your chat conversations as Excel files. This is useful for:
@@ -23,6 +36,13 @@ The Export to Excel plugin allows you to download your chat conversations as Exc
 - :material-download: **One-Click Download**: Instant file generation
 - :material-history: **Full History**: Exports complete conversation
 
+## Configuration
+
+- **Title Source**: Choose how the filename is generated:
+  - `chat_title`: Use the chat title (default).
+  - `ai_generated`: Use AI to generate a concise title from the content.
+  - `markdown_title`: Extract the first H1/H2 header from the markdown content.
+
 ---
 
 ## Installation

docs/plugins/actions/export-to-excel.zh.md

@@ -1,12 +1,25 @@
 # Export to Excel（导出到 Excel）
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v1.0.0</span>
+<span class="version-badge">v0.3.7</span>
 
 将聊天记录导出为 Excel 表格，便于分析、归档和分享。
 
+
+### v0.3.6 更新内容
+- **OpenWebUI 风格主题**：现代深灰表头，搭配浅灰斑马纹，提升可读性。
+- **斑马纹效果**：隔行变色，方便视觉扫描。
+- **智能数据类型转换**：自动将列转换为数字或日期类型。
+- **全单元格粗体/斜体**：支持 Markdown 粗体/斜体格式。
+- **部分 Markdown 清理**：移除部分 Markdown 符号，输出更整洁。
+- **导出范围**：可选择导出"最后一条消息"或"所有消息"。
+- **智能 Sheet 命名**：根据 Markdown 标题或消息索引命名 Sheet。
+- **智能文件名生成**：支持对话标题、AI 总结或 Markdown 标题生成文件名。
+- **AI 标题生成**：支持指定模型 (`MODEL_ID`) 生成标题，并提供生成进度通知。
+
 ---
 
+
 ## 概览
 
 Export to Excel 插件可以把你的聊天记录下载为 Excel 文件，适用于：
@@ -23,6 +36,13 @@ Export to Excel 插件可以把你的聊天记录下载为 Excel 文件，适用
 - :material-download: **一键下载**：即时生成文件
 - :material-history: **完整历史**：导出完整会话内容
 
+## 配置
+
+- **标题来源 (Title Source)**：选择文件名的生成方式：
+  - `chat_title`：使用对话标题（默认）。
+  - `ai_generated`：使用 AI 根据内容生成简洁标题。
+  - `markdown_title`：提取 Markdown 内容中的第一个 H1/H2 标题。
+
 ---
 
 ## 安装

docs/plugins/actions/export-to-word.md

@@ -1,9 +1,9 @@
 # Export to Word
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v0.1.0</span>
+<span class="version-badge">v0.4.3</span>
 
-Export chat conversations to Word (.docx) with Markdown formatting, syntax highlighting, and smarter filenames.
+Export conversation to Word (.docx) with **syntax highlighting**, **native math equations**, **Mermaid diagrams**, **citations**, and **enhanced table formatting**.
 
 ---
 
@@ -13,11 +13,17 @@ The Export to Word plugin converts chat messages from Markdown to a polished Wor
 
 ## Features
 
-- :material-file-word-box: **DOCX Export**: Generate Word files with one click
-- :material-format-bold: **Rich Markdown Support**: Headings, bold/italic, lists, tables
-- :material-code-tags: **Syntax Highlighting**: Pygments-powered code blocks
-- :material-format-quote-close: **Styled Blockquotes**: Left-border gray quote styling
-- :material-file-document-outline: **Smart Filenames**: Configurable title source (Chat Title, AI Generated, or Markdown Title)
+- :material-file-word-box: **One-Click Export**: Adds an "Export to Word" action button to the chat.
+- :material-format-bold: **Markdown Conversion**: Converts Markdown syntax to Word formatting (headings, bold, italic, code, tables, lists).
+- :material-code-tags: **Syntax Highlighting**: Code blocks are highlighted with Pygments (supports 500+ languages).
+- :material-sigma: **Native Math Equations**: LaTeX math (`$$...$$`, `\[...\]`, `$...$`, `\(...\)`) converted to editable Word equations.
+- :material-graph: **Mermaid Diagrams**: Mermaid flowcharts and sequence diagrams rendered as images in the document.
+- :material-book-open-page-variant: **Citations & References**: Auto-generates a References section from OpenWebUI sources with clickable citation links.
+- :material-brain-off: **Reasoning Stripping**: Automatically removes AI thinking blocks (`<think>`, `<analysis>`) from exports.
+- :material-table: **Enhanced Tables**: Smart column widths, column alignment (`:---`, `---:`, `:---:`), header row repeat across pages.
+- :material-format-quote-close: **Blockquote Support**: Markdown blockquotes are rendered with left border and gray styling.
+- :material-translate: **Multi-language Support**: Properly handles both Chinese and English text.
+- :material-file-document-outline: **Smarter Filenames**: Configurable title source (Chat Title, AI Generated, or Markdown Title).
 
 ---
 
@@ -26,8 +32,38 @@ The Export to Word plugin converts chat messages from Markdown to a polished Wor
 You can configure the following settings via the **Valves** button in the plugin settings:
 
 | Valve | Description | Default |
-| :------------- | :------------------------------------------------------------------------------------------ | :----------- |
+| :--- | :--- | :--- |
 | `TITLE_SOURCE` | Source for document title/filename. Options: `chat_title`, `ai_generated`, `markdown_title` | `chat_title` |
+| `MAX_EMBED_IMAGE_MB` | Maximum image size to embed into DOCX (MB). | `20` |
+| `UI_LANGUAGE` | User interface language. Options: `en` (English), `zh` (Chinese). | `en` |
+| `FONT_LATIN` | Font name for Latin characters. | `Times New Roman` |
+| `FONT_ASIAN` | Font name for Asian characters. | `SimSun` |
+| `FONT_CODE` | Font name for code blocks. | `Consolas` |
+| `TABLE_HEADER_COLOR` | Table header background color (Hex without #). | `F2F2F2` |
+| `TABLE_ZEBRA_COLOR` | Table alternating row background color (Hex without #). | `FBFBFB` |
+| `MERMAID_JS_URL` | URL for the Mermaid.js library. | `https://cdn.jsdelivr.net/npm/mermaid@11.12.2/dist/mermaid.min.js` |
+| `MERMAID_JSZIP_URL` | URL for the JSZip library (required for DOCX manipulation). | `https://cdnjs.cloudflare.com/ajax/libs/jszip/3.10.1/jszip.min.js` |
+| `MERMAID_PNG_SCALE` | Scale factor for Mermaid PNG generation (Resolution). | `3.0` |
+| `MERMAID_DISPLAY_SCALE` | Scale factor for Mermaid visual size in Word. | `1.0` |
+| `MERMAID_OPTIMIZE_LAYOUT` | Automatically convert LR (Left-Right) flowcharts to TD (Top-Down). | `False` |
+| `MERMAID_BACKGROUND` | Background color for Mermaid diagrams (e.g., `white`, `transparent`). | `transparent` |
+| `MERMAID_CAPTIONS_ENABLE` | Enable/disable figure captions for Mermaid diagrams. | `True` |
+| `MERMAID_CAPTION_STYLE` | Paragraph style name for Mermaid captions. | `Caption` |
+| `MERMAID_CAPTION_PREFIX` | Caption prefix label (e.g., 'Figure'). Empty = auto-detect based on language. | `""` |
+| `MATH_ENABLE` | Enable LaTeX math block conversion. | `True` |
+| `MATH_INLINE_DOLLAR_ENABLE` | Enable inline `$ ... $` math conversion. | `True` |
+
+## 🔥 What's New in v0.4.3
+
+### User-Level Configuration (UserValves)
+
+Users can override the following settings in their personal settings:
+- `TITLE_SOURCE`
+- `UI_LANGUAGE`
+- `FONT_LATIN`, `FONT_ASIAN`, `FONT_CODE`
+- `TABLE_HEADER_COLOR`, `TABLE_ZEBRA_COLOR`
+- `MERMAID_...` (Selected Mermaid settings)
+- `MATH_...` (Math settings)
 
 ---
 
@@ -47,34 +83,41 @@ You can configure the following settings via the **Valves** button in the plugin
 
 ---
 
-## Supported Markdown
+## Supported Markdown Syntax
 
 | Syntax | Word Result |
-| :---------------------------------- | :----------------------------- |
+| :--- | :--- |
 | `# Heading 1` to `###### Heading 6` | Heading levels 1-6 |
-| `**bold**` / `__bold__`             | Bold text                      |
-| `*italic*` / `_italic_`             | Italic text                    |
+| `**bold**` or `__bold__` | Bold text |
+| `*italic*` or `_italic_` | Italic text |
 | `***bold italic***` | Bold + Italic |
 | `` `inline code` `` | Monospace with gray background |
-| <code>``` code block ```</code>     | Syntax-highlighted code block  |
+| ` ``` code block ``` ` | **Syntax highlighted** code block |
 | `> blockquote` | Left-bordered gray italic text |
-| `[link](url)`                       | Blue underlined link           |
-| `~~strikethrough~~`                 | Strikethrough                  |
-| `- item` / `* item`                 | Bullet list                    |
+| `[link](url)` | Blue underlined link text |
+| `~~strikethrough~~` | Strikethrough text |
+| `- item` or `* item` | Bullet list |
 | `1. item` | Numbered list |
-| Markdown tables                     | Grid table                     |
-| `---` / `***`                       | Horizontal rule                |
+| Markdown tables | **Enhanced table** with smart widths |
+| `---` or `***` | Horizontal rule |
+| `$$LaTeX$$` or `\[LaTeX\]` | **Native Word equation** (display) |
+| `$LaTeX$` or `\(LaTeX\)` | **Native Word equation** (inline) |
+| ` ```mermaid ... ``` ` | **Mermaid diagram** as image |
+| `[1]` citation markers | **Clickable links** to References |
 
 ---
 
 ## Requirements
 
 !!! note "Prerequisites"
-    - `python-docx==1.1.2` (document generation)
-    - `Pygments>=2.15.0` (syntax highlighting, optional but recommended)
+    - `python-docx==1.1.2` - Word document generation
+    - `Pygments>=2.15.0` - Syntax highlighting
+    - `latex2mathml` - LaTeX to MathML conversion
+    - `mathml2omml` - MathML to Office Math (OMML) conversion
 
 ---
 
 ## Source Code
 
 [:fontawesome-brands-github: View on GitHub](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/export_to_docx){ .md-button }
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.4.3 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)

docs/plugins/actions/export-to-word.zh.md

@@ -1,9 +1,9 @@
 # Export to Word（导出为 Word）
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v0.1.0</span>
+<span class="version-badge">v0.4.3</span>
 
-将聊天记录按 Markdown 格式导出为 Word (.docx)，支持语法高亮、引用样式和更智能的文件命名。
+将当前对话导出为完美格式的 Word 文档，支持**代码语法高亮**、**原生数学公式**、**Mermaid 图表**、**引用资料**以及**增强表格**渲染。
 
 ---
 
@@ -13,11 +13,17 @@ Export to Word 插件会把聊天消息从 Markdown 转成精致的 Word 文档
 
 ## 功能特性
 
-- :material-file-word-box: **DOCX 导出**：一键生成 Word 文件
-- :material-format-bold: **丰富 Markdown 支持**：标题、粗斜体、列表、表格
-- :material-code-tags: **语法高亮**：Pygments 驱动的代码块上色
-- :material-format-quote-close: **引用样式**：左侧边框的灰色斜体引用
-- :material-file-document-outline: **智能文件名**：可配置标题来源（对话标题、AI 生成或 Markdown 标题）
+- :material-file-word-box: **一键导出**：在聊天界面添加"导出为 Word"动作按钮。
+- :material-format-bold: **Markdown 转换**：将 Markdown 语法转换为 Word 格式（标题、粗体、斜体、代码、表格、列表）。
+- :material-code-tags: **代码语法高亮**：使用 Pygments 库为代码块添加语法高亮（支持 500+ 种语言）。
+- :material-sigma: **原生数学公式**：LaTeX 公式（`$$...$$`、`\[...\]`、`$...$`、`\(...\)`）转换为可编辑的 Word 公式。
+- :material-graph: **Mermaid 图表**：Mermaid 流程图和时序图渲染为文档中的图片。
+- :material-book-open-page-variant: **引用与参考**：自动从 OpenWebUI 来源生成参考资料章节，支持可点击的引用链接。
+- :material-brain-off: **移除思考过程**：自动移除 AI 思考块（`<think>`、`<analysis>`）。
+- :material-table: **增强表格**：智能列宽、列对齐（`:---`、`---:`、`:---:`）、表头跨页重复。
+- :material-format-quote-close: **引用块支持**：Markdown 引用块渲染为带左侧边框的灰色斜体样式。
+- :material-translate: **多语言支持**：正确处理中文和英文文本，无乱码问题。
+- :material-file-document-outline: **智能文件名**：可配置标题来源（对话标题、AI 生成或 Markdown 标题）。
 
 ---
 
@@ -26,8 +32,36 @@ Export to Word 插件会把聊天消息从 Markdown 转成精致的 Word 文档
 您可以通过插件设置中的 **Valves** 按钮配置以下选项：
 
 | Valve | 说明 | 默认值 |
-| :------------- | :--------------------------------------------------------------------------------------------------------------- | :----------- |
-| `TITLE_SOURCE` | 文档标题/文件名的来源。选项：`chat_title` (对话标题), `ai_generated` (AI 生成), `markdown_title` (Markdown 标题) | `chat_title` |
+| :--- | :--- | :--- |
+| `文档标题来源` | 文档标题/文件名的来源。选项：`chat_title` (对话标题), `ai_generated` (AI 生成), `markdown_title` (Markdown 标题) | `chat_title` |
+| `最大嵌入图片大小MB` | 嵌入图片的最大大小 (MB)。 | `20` |
+| `界面语言` | 界面语言。选项：`en` (英语), `zh` (中文)。 | `zh` |
+| `英文字体` | 英文字体名称。 | `Calibri` |
+| `中文字体` | 中文字体名称。 | `SimSun` |
+| `代码字体` | 代码字体名称。 | `Consolas` |
+| `表头背景色` | 表头背景色（十六进制，不带#）。 | `F2F2F2` |
+| `表格隔行背景色` | 表格隔行背景色（十六进制，不带#）。 | `FBFBFB` |
+| `Mermaid_JS地址` | Mermaid.js 库的 URL。 | `https://cdn.jsdelivr.net/npm/mermaid@11.12.2/dist/mermaid.min.js` |
+| `JSZip库地址` | JSZip 库的 URL（用于 DOCX 操作）。 | `https://cdnjs.cloudflare.com/ajax/libs/jszip/3.10.1/jszip.min.js` |
+| `Mermaid_PNG缩放比例` | Mermaid PNG 生成缩放比例（分辨率）。 | `3.0` |
+| `Mermaid显示比例` | Mermaid 在 Word 中的显示比例（视觉大小）。 | `1.0` |
+| `Mermaid布局优化` | 优化 Mermaid 布局: 自动将 LR (左右) 转换为 TD (上下)。 | `False` |
+| `Mermaid背景色` | Mermaid 图表背景色（如 `white`, `transparent`）。 | `transparent` |
+| `启用Mermaid图注` | 启用/禁用 Mermaid 图表的图注。 | `True` |
+| `Mermaid图注样式` | Mermaid 图注的段落样式名称。 | `Caption` |
+| `Mermaid图注前缀` | 图注前缀（如 '图'）。留空则根据语言自动检测。 | `""` |
+| `启用数学公式` | 启用 LaTeX 数学公式块转换。 | `True` |
+| `启用行内公式` | 启用行内 `$ ... $` 数学公式转换。 | `True` |
+
+### 用户级配置 (UserValves)
+
+用户可以在个人设置中覆盖以下配置：
+- `文档标题来源`
+- `界面语言`
+- `英文字体`, `中文字体`, `代码字体`
+- `表头背景色`, `表格隔行背景色`
+- `Mermaid_...` (部分 Mermaid 设置)
+- `启用数学公式`, `启用行内公式`
 
 ---
 
@@ -47,10 +81,10 @@ Export to Word 插件会把聊天消息从 Markdown 转成精致的 Word 文档
 
 ---
 
-## 支持的 Markdown
+## 支持的 Markdown 语法
 
 | 语法 | Word 效果 |
-| :-------------------------- | :------------------ |
+| :--- | :--- |
 | `# 标题1` 到 `###### 标题6` | 标题级别 1-6 |
 | `**粗体**` / `__粗体__` | 粗体文本 |
 | `*斜体*` / `_斜体_` | 斜体文本 |
@@ -62,8 +96,12 @@ Export to Word 插件会把聊天消息从 Markdown 转成精致的 Word 文档
 | `~~删除线~~` | 删除线 |
 | `- 项目` / `* 项目` | 无序列表 |
 | `1. 项目` | 有序列表 |
-| Markdown 表格               | 带边框表格          |
+| Markdown 表格 | **增强表格**（智能列宽） |
 | `---` / `***` | 水平分割线 |
+| `$$LaTeX$$` 或 `\[LaTeX\]` | **原生 Word 公式**（块级） |
+| `$LaTeX$` 或 `\(LaTeX\)` | **原生 Word 公式**（行内） |
+| ` ```mermaid ... ``` ` | **Mermaid 图表**（图片形式） |
+| `[1]` 引用标记 | **可点击链接**到参考资料 |
 
 ---
 
@@ -71,10 +109,12 @@ Export to Word 插件会把聊天消息从 Markdown 转成精致的 Word 文档
 
 !!! note "前置条件"
     - `python-docx==1.1.2`（文档生成）
-    - `Pygments>=2.15.0`（语法高亮，建议安装）
+    - `Pygments>=2.15.0`（语法高亮）
+    - `latex2mathml`（LaTeX 转 MathML）
+    - `mathml2omml`（MathML 转 Office Math）
 
 ---
 
 ## 源码
 
-[:fontawesome-brands-github: 在 GitHub 查看](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/export_to_docx){ .md-button }
+[:fontawes**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.4.3 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)/tree/main/plugins/actions/export_to_docx){ .md-button }

docs/plugins/actions/index.md

@@ -33,7 +33,7 @@ Actions are interactive plugins that:
 
     Transform text into professional infographics using AntV visualization engine with various templates.
 
-    **Version:** 1.3.0
+    **Version:** 1.4.9
 
     [:octicons-arrow-right-24: Documentation](smart-infographic.md)
 
@@ -53,29 +53,39 @@ Actions are interactive plugins that:
 
     Export chat conversations to Excel spreadsheet format for analysis and archiving.
 
-    **Version:** 0.3.3
+    **Version:** 0.3.7
 
     [:octicons-arrow-right-24: Documentation](export-to-excel.md)
 
--   :material-file-word-box:{ .lg .middle } **Export to Word**
+-   :material-file-word-box:{ .lg .middle } **Export to Word (Enhanced Formatting)**
 
     ---
 
-    Export chat content as Word (.docx) with Markdown formatting and syntax highlighting.
+    Export the current conversation to a formatted Word doc with **syntax highlighting**, **native math equations**, **Mermaid diagrams**, **citations**, and **enhanced table formatting**.
     
-    **Version:** 0.1.0
+    **Version:** 0.4.2
 
     [:octicons-arrow-right-24: Documentation](export-to-word.md)
 
--   :material-text-box-search:{ .lg .middle } **Summary**
+-   :material-brain:{ .lg .middle } **Deep Dive**
 
     ---
 
-    Generate concise summaries of long text content with key points extraction.
+    A comprehensive thinking lens that dives deep into any content - Context → Logic → Insight → Path. Supports theme auto-adaptation.
 
-    **Version:** 0.1.0
+    **Version:** 1.0.0
 
-    [:octicons-arrow-right-24: Documentation](summary.md)
+    [:octicons-arrow-right-24: Documentation](deep-dive.md)
+
+-   :material-image-text:{ .lg .middle } **Infographic to Markdown**
+
+    ---
+
+    AI-powered infographic generator that renders SVG and embeds it as Markdown Data URL image.
+
+    **Version:** 1.0.0
+
+    [:octicons-arrow-right-24: Documentation](infographic-markdown.md)
 
 </div>
 

docs/plugins/actions/index.zh.md

@@ -33,7 +33,7 @@ Actions 是交互式插件，能够：
 
     使用 AntV 可视化引擎，将文本转成专业的信息图。
 
-    **版本：** 1.3.0
+    **版本：** 1.4.9
 
     [:octicons-arrow-right-24: 查看文档](smart-infographic.md)
 
@@ -53,29 +53,39 @@ Actions 是交互式插件，能够：
 
     将聊天记录导出为 Excel 电子表格，方便分析或归档。
 
-    **版本：** 0.3.3
+    **版本：** 0.3.7
 
     [:octicons-arrow-right-24: 查看文档](export-to-excel.md)
 
--   :material-file-word-box:{ .lg .middle } **Export to Word**
+-   :material-file-word-box:{ .lg .middle } **Word 导出 (格式增强)**
 
     ---
 
-    将聊天内容按 Markdown 格式导出为 Word (.docx)，支持语法高亮。
+    将当前对话导出为完美格式的 Word 文档，支持**代码语法高亮**、**原生数学公式**、**Mermaid 图表**、**引用资料**以及**增强表格**渲染。
 
-    **版本：** 0.1.0
+    **版本：** 0.4.2
 
     [:octicons-arrow-right-24: 查看文档](export-to-word.md)
 
--   :material-text-box-search:{ .lg .middle } **Summary**
+-   :material-brain:{ .lg .middle } **精读 (Deep Dive)**
 
     ---
 
-    对长文本进行精简总结，提取要点。
+    全方位的思维透镜 —— 全景 → 脉络 → 洞察 → 路径。支持主题自适应。
 
-    **版本：** 0.1.0
+    **版本：** 1.0.0
 
-    [:octicons-arrow-right-24: 查看文档](summary.md)
+    [:octicons-arrow-right-24: 查看文档](deep-dive.zh.md)
+
+-   :material-image-text:{ .lg .middle } **信息图转 Markdown**
+
+    ---
+
+    AI 驱动的信息图生成器，渲染 SVG 并以 Markdown Data URL 图片嵌入。
+
+    **版本：** 1.0.0
+
+    [:octicons-arrow-right-24: 查看文档](infographic-markdown.zh.md)
 
 </div>
 

docs/plugins/actions/infographic-markdown.md

@@ -0,0 +1,120 @@
+# Infographic to Markdown
+
+> **Version:** 1.0.0 | **Author:** Fu-Jie
+
+AI-powered infographic generator that renders SVG on the frontend and embeds it directly into Markdown as a Data URL image.
+
+## Overview
+
+This plugin combines the power of AI text analysis with AntV Infographic visualization to create beautiful infographics that are embedded directly into chat messages as Markdown images.
+
+### Key Features
+
+- :robot: **AI-Powered**: Automatically analyzes text and selects the best infographic template
+- :bar_chart: **Multiple Templates**: Supports 18+ infographic templates (lists, charts, comparisons, etc.)
+- :framed_picture: **Self-Contained**: SVG/PNG embedded as Data URL, no external dependencies
+- :memo: **Markdown Native**: Results are pure Markdown images, compatible everywhere
+- :arrows_counterclockwise: **API Writeback**: Updates message content via REST API for persistence
+
+### How It Works
+
+```mermaid
+graph TD
+    A[User triggers action] --> B[Python extracts message content]
+    B --> C[LLM generates Infographic syntax]
+    C --> D[Frontend JS loads AntV library]
+    D --> E[Render SVG offscreen]
+    E --> F[Export to Data URL]
+    F --> G[Update message via API]
+    G --> H[Display as Markdown image]
+```
+
+## Installation
+
+1. Download `infographic_markdown.py` (English) or `infographic_markdown_cn.py` (Chinese)
+2. Navigate to **Admin Panel** → **Settings** → **Functions**
+3. Upload the file and configure settings
+4. Use the action button in chat messages
+
+## Configuration
+
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| `SHOW_STATUS` | bool | `true` | Show operation status updates |
+| `MODEL_ID` | string | `""` | LLM model ID (empty = use current model) |
+| `MIN_TEXT_LENGTH` | int | `50` | Minimum text length required |
+| `MESSAGE_COUNT` | int | `1` | Number of recent messages to use |
+| `SVG_WIDTH` | int | `800` | Width of generated SVG (pixels) |
+| `EXPORT_FORMAT` | string | `"svg"` | Export format: `svg` or `png` |
+
+## Supported Templates
+
+| Category | Template | Description |
+|----------|----------|-------------|
+| List | `list-grid` | Grid cards |
+| List | `list-vertical` | Vertical list |
+| Tree | `tree-vertical` | Vertical tree |
+| Tree | `tree-horizontal` | Horizontal tree |
+| Mind Map | `mindmap` | Mind map |
+| Process | `sequence-roadmap` | Roadmap |
+| Process | `sequence-zigzag` | Zigzag process |
+| Relation | `relation-sankey` | Sankey diagram |
+| Relation | `relation-circle` | Circular relation |
+| Compare | `compare-binary` | Binary comparison |
+| Analysis | `compare-swot` | SWOT analysis |
+| Quadrant | `quadrant-quarter` | Quadrant chart |
+| Chart | `chart-bar` | Bar chart |
+| Chart | `chart-column` | Column chart |
+| Chart | `chart-line` | Line chart |
+| Chart | `chart-pie` | Pie chart |
+| Chart | `chart-doughnut` | Doughnut chart |
+| Chart | `chart-area` | Area chart |
+
+## Usage Example
+
+1. Generate some text content in the chat (or have the AI generate it)
+2. Click the **📊 Infographic to Markdown** action button
+3. Wait for AI analysis and SVG rendering
+4. The infographic will be embedded as a Markdown image
+
+## Technical Details
+
+### Data URL Embedding
+
+The plugin converts SVG graphics to Base64-encoded Data URLs:
+
+```javascript
+const svgData = new XMLSerializer().serializeToString(svg);
+const base64 = btoa(unescape(encodeURIComponent(svgData)));
+const dataUri = "data:image/svg+xml;base64," + base64;
+const markdownImage = `![description](${dataUri})`;
+```
+
+### AntV toDataURL API
+
+```javascript
+// Export as SVG (recommended)
+const svgUrl = await instance.toDataURL({
+    type: 'svg',
+    embedResources: true
+});
+
+// Export as PNG
+const pngUrl = await instance.toDataURL({
+    type: 'png',
+    dpr: 2
+});
+```
+
+## Notes
+
+1. **Browser Compatibility**: Requires modern browsers with ES6+ and Fetch API support
+2. **Network Dependency**: First use requires loading AntV library from CDN
+3. **Data URL Size**: Base64 encoding increases size by ~33%
+4. **Chinese Fonts**: SVG export embeds fonts for correct display
+
+## Related Resources
+
+- [AntV Infographic Documentation](https://infographic.antv.vision/)
+- [Infographic API Reference](https://infographic.antv.vision/reference/infographic-api)
+- [Infographic Syntax Guide](https://infographic.antv.vision/learn/infographic-syntax)

docs/plugins/actions/infographic-markdown.zh.md

@@ -0,0 +1,120 @@
+# 信息图转 Markdown
+
+> **版本:** 1.0.0 | **作者:** Fu-Jie
+
+AI 驱动的信息图生成器，在前端渲染 SVG 并以 Data URL 图片格式直接嵌入到 Markdown 中。
+
+## 概述
+
+这个插件结合了 AI 文本分析能力和 AntV Infographic 可视化引擎，生成精美的信息图并以 Markdown 图片格式直接嵌入到聊天消息中。
+
+### 主要特性
+
+- :robot: **AI 驱动**: 自动分析文本并选择最佳的信息图模板
+- :bar_chart: **多种模板**: 支持 18+ 种信息图模板（列表、图表、对比等）
+- :framed_picture: **自包含**: SVG/PNG 以 Data URL 嵌入，无外部依赖
+- :memo: **Markdown 原生**: 结果是纯 Markdown 图片，兼容任何平台
+- :arrows_counterclockwise: **API 回写**: 通过 REST API 更新消息内容实现持久化
+
+### 工作原理
+
+```mermaid
+graph TD
+    A[用户触发动作] --> B[Python 提取消息内容]
+    B --> C[LLM 生成 Infographic 语法]
+    C --> D[前端 JS 加载 AntV 库]
+    D --> E[离屏渲染 SVG]
+    E --> F[导出为 Data URL]
+    F --> G[通过 API 更新消息]
+    G --> H[显示为 Markdown 图片]
+```
+
+## 安装
+
+1. 下载 `infographic_markdown.py`（英文版）或 `infographic_markdown_cn.py`（中文版）
+2. 进入 **管理面板** → **设置** → **功能**
+3. 上传文件并配置设置
+4. 在聊天消息中使用动作按钮
+
+## 配置选项
+
+| 参数 | 类型 | 默认值 | 描述 |
+|------|------|--------|------|
+| `SHOW_STATUS` | bool | `true` | 是否显示操作状态 |
+| `MODEL_ID` | string | `""` | LLM 模型 ID（空则使用当前模型） |
+| `MIN_TEXT_LENGTH` | int | `50` | 最小文本长度要求 |
+| `MESSAGE_COUNT` | int | `1` | 用于生成的最近消息数量 |
+| `SVG_WIDTH` | int | `800` | 生成的 SVG 宽度（像素） |
+| `EXPORT_FORMAT` | string | `"svg"` | 导出格式：`svg` 或 `png` |
+
+## 支持的模板
+
+| 类别 | 模板名称 | 描述 |
+|------|----------|------|
+| 列表 | `list-grid` | 网格卡片 |
+| 列表 | `list-vertical` | 垂直列表 |
+| 树形 | `tree-vertical` | 垂直树 |
+| 树形 | `tree-horizontal` | 水平树 |
+| 思维导图 | `mindmap` | 思维导图 |
+| 流程 | `sequence-roadmap` | 路线图 |
+| 流程 | `sequence-zigzag` | 折线流程 |
+| 关系 | `relation-sankey` | 桑基图 |
+| 关系 | `relation-circle` | 圆形关系 |
+| 对比 | `compare-binary` | 二元对比 |
+| 分析 | `compare-swot` | SWOT 分析 |
+| 象限 | `quadrant-quarter` | 四象限图 |
+| 图表 | `chart-bar` | 条形图 |
+| 图表 | `chart-column` | 柱状图 |
+| 图表 | `chart-line` | 折线图 |
+| 图表 | `chart-pie` | 饼图 |
+| 图表 | `chart-doughnut` | 环形图 |
+| 图表 | `chart-area` | 面积图 |
+
+## 使用示例
+
+1. 在聊天中生成一些文本内容（或让 AI 生成）
+2. 点击 **📊 信息图转 Markdown** 动作按钮
+3. 等待 AI 分析和 SVG 渲染
+4. 信息图将以 Markdown 图片形式嵌入
+
+## 技术细节
+
+### Data URL 嵌入
+
+插件将 SVG 图形转换为 Base64 编码的 Data URL：
+
+```javascript
+const svgData = new XMLSerializer().serializeToString(svg);
+const base64 = btoa(unescape(encodeURIComponent(svgData)));
+const dataUri = "data:image/svg+xml;base64," + base64;
+const markdownImage = `![描述](${dataUri})`;
+```
+
+### AntV toDataURL API
+
+```javascript
+// 导出 SVG（推荐）
+const svgUrl = await instance.toDataURL({
+    type: 'svg',
+    embedResources: true
+});
+
+// 导出 PNG
+const pngUrl = await instance.toDataURL({
+    type: 'png',
+    dpr: 2
+});
+```
+
+## 注意事项
+
+1. **浏览器兼容性**: 需要现代浏览器支持 ES6+ 和 Fetch API
+2. **网络依赖**: 首次使用需要从 CDN 加载 AntV Infographic 库
+3. **Data URL 大小**: Base64 编码会增加约 33% 的体积
+4. **中文字体**: SVG 导出时会嵌入字体以确保正确显示
+
+## 相关资源
+
+- [AntV Infographic 官方文档](https://infographic.antv.vision/)
+- [Infographic API 参考](https://infographic.antv.vision/reference/infographic-api)
+- [Infographic 语法规范](https://infographic.antv.vision/learn/infographic-syntax)

docs/plugins/actions/smart-infographic.md

@@ -1,7 +1,7 @@
 # Smart Infographic
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v1.3.0</span>
+<span class="version-badge">v1.4.9</span>
 
 An AntV Infographic engine powered plugin that transforms long text into professional, beautiful infographics with a single click.
 
@@ -14,11 +14,13 @@ The Smart Infographic plugin uses AI to analyze text content and generate profes
 ## Features
 
 - :material-robot: **AI-Powered Transformation**: Automatically analyzes text logic, extracts key points, and generates structured charts
-- :material-palette: **Professional Templates**: Includes various AntV official templates: Lists, Trees, Mindmaps, Comparison Tables, Flowcharts, and Statistical Charts
-- :material-magnify: **Auto-Icon Matching**: Built-in logic to search and match the most relevant Material Design Icons based on content
+- :material-palette: **70+ Professional Templates**: Includes various AntV official templates: Lists, Trees, Roadmaps, Timelines, Comparison Tables, SWOT, Quadrants, and Statistical Charts
+- :material-magnify: **Auto-Icon Matching**: Built-in logic to search and match the most relevant icons (Iconify) and illustrations (unDraw)
 - :material-download: **Multi-Format Export**: Download your infographics as **SVG**, **PNG**, or **Standalone HTML** file
 - :material-theme-light-dark: **Theme Support**: Supports Dark/Light modes, auto-adapts theme colors
 - :material-cellphone-link: **Responsive Design**: Generated charts look great on both desktop and mobile devices
+- :material-image: **Image Embedding**: Option to embed charts as static images for better compatibility
+- :material-monitor-screenshot: **Adaptive Sizing**: Images automatically adapt to the chat container width
 
 ---
 
@@ -35,10 +37,11 @@ The Smart Infographic plugin uses AI to analyze text content and generate profes
 
 | Category | Template Name | Use Case |
 |:---------|:--------------|:---------|
-| **Lists & Hierarchy** | `list-grid`, `tree-vertical`, `mindmap` | Features, Org Charts, Brainstorming |
-| **Sequence & Relation** | `sequence-roadmap`, `relation-circle` | Roadmaps, Circular Flows, Steps |
-| **Comparison & Analysis** | `compare-binary`, `compare-swot`, `quadrant-quarter` | Pros/Cons, SWOT, Quadrants |
-| **Charts & Data** | `chart-bar`, `chart-line`, `chart-pie` | Trends, Distributions, Metrics |
+| **Sequence** | `sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-snake-steps-compact-card` | Timelines, Roadmaps, Processes |
+| **Lists** | `list-grid-candy-card-lite`, `list-row-horizontal-icon-arrow`, `list-column-simple-vertical-arrow` | Features, Bullet Points, Lists |
+| **Comparison** | `compare-binary-horizontal-underline-text-vs`, `compare-swot`, `quadrant-quarter-simple-card` | Pros/Cons, SWOT, Quadrants |
+| **Hierarchy** | `hierarchy-tree-tech-style-capsule-item`, `hierarchy-structure` | Org Charts, Structures |
+| **Charts** | `chart-column-simple`, `chart-bar-plain-text`, `chart-line-plain-text`, `chart-wordcloud` | Trends, Distributions, Metrics |
 
 ---
 
@@ -60,6 +63,7 @@ The Smart Infographic plugin uses AI to analyze text content and generate profes
 | `MIN_TEXT_LENGTH` | integer | `100` | Minimum characters required to trigger analysis |
 | `CLEAR_PREVIOUS_HTML` | boolean | `false` | Whether to clear previous charts |
 | `MESSAGE_COUNT` | integer | `1` | Number of recent messages to use for analysis |
+| `OUTPUT_MODE` | string | `image` | `image` for static image embedding (default), `html` for interactive chart |
 
 ---
 

docs/plugins/actions/smart-infographic.zh.md

@@ -1,7 +1,7 @@
 # Smart Infographic（智能信息图）
 
 <span class="category-badge action">Action</span>
-<span class="version-badge">v1.0.0</span>
+<span class="version-badge">v1.4.9</span>
 
 基于 AntV 信息图引擎，将长文本一键转成专业、美观的信息图。
 
@@ -14,11 +14,13 @@ Smart Infographic 使用 AI 分析文本，并基于 AntV 可视化引擎生成
 ## 功能特性
 
 - :material-robot: **AI 转换**：自动分析文本逻辑，提取要点并生成结构化图表
-- :material-palette: **专业模板**：内置 AntV 官方模板：列表、树、思维导图、对比表、流程图、统计图等
-- :material-magnify: **自动匹配图标**：根据内容自动选择最合适的 Material Design Icons
+- :material-palette: **70+ 专业模板**：内置多种 AntV 官方模板，包括列表、树图、路线图、时间线、对比图、SWOT、象限图及统计图表等
+- :material-magnify: **自动匹配图标**：内置图标搜索逻辑，支持 Iconify 图标和 unDraw 插图自动匹配
 - :material-download: **多格式导出**：支持下载 **SVG**、**PNG**、**独立 HTML**
 - :material-theme-light-dark: **主题支持**：适配深色/浅色模式
 - :material-cellphone-link: **响应式**：桌面与移动端都能良好展示
+- :material-image: **图片嵌入**：支持将图表作为静态图片嵌入，兼容性更好
+- :material-monitor-screenshot: **自适应尺寸**：图片模式下自动适应聊天容器宽度
 
 ---
 
@@ -35,10 +37,11 @@ Smart Infographic 使用 AI 分析文本，并基于 AntV 可视化引擎生成
 
 | 分类 | 模板名称 | 典型场景 |
 |:---------|:--------------|:---------|
-| **列表与层级** | `list-grid`, `tree-vertical`, `mindmap` | 特性列表、组织结构、头脑风暴 |
-| **序列与关系** | `sequence-roadmap`, `relation-circle` | 路线图、循环流程、步骤拆解 |
-| **对比与分析** | `compare-binary`, `compare-swot`, `quadrant-quarter` | 优劣势、SWOT、象限分析 |
-| **图表与数据** | `chart-bar`, `chart-line`, `chart-pie` | 趋势、分布、指标对比 |
+| **时序与流程** | `sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-snake-steps-compact-card` | 时间线、路线图、步骤说明 |
+| **列表与网格** | `list-grid-candy-card-lite`, `list-row-horizontal-icon-arrow`, `list-column-simple-vertical-arrow` | 功能亮点、要点列举、清单 |
+| **对比与分析** | `compare-binary-horizontal-underline-text-vs`, `compare-swot`, `quadrant-quarter-simple-card` | 优劣势对比、SWOT 分析、象限图 |
+| **层级与结构** | `hierarchy-tree-tech-style-capsule-item`, `hierarchy-structure` | 组织架构、层级关系 |
+| **图表与数据** | `chart-column-simple`, `chart-bar-plain-text`, `chart-line-plain-text`, `chart-wordcloud` | 数据趋势、比例分布、数值对比 |
 
 ---
 
@@ -60,6 +63,7 @@ Smart Infographic 使用 AI 分析文本，并基于 AntV 可视化引擎生成
 | `MIN_TEXT_LENGTH` | integer | `100` | 触发分析的最小字符数 |
 | `CLEAR_PREVIOUS_HTML` | boolean | `false` | 是否清空之前生成的图表 |
 | `MESSAGE_COUNT` | integer | `1` | 参与分析的最近消息条数 |
+| `OUTPUT_MODE` | string | `image` | `image` 为静态图片嵌入（默认），`html` 为交互式图表 |
 
 ---
 

docs/plugins/actions/summary.md

@@ -1,82 +0,0 @@
-# Summary
-
-<span class="category-badge action">Action</span>
-<span class="version-badge">v0.1.0</span>
-
-Generate concise summaries of long text content with key points extraction.
-
----
-
-## Overview
-
-The Summary plugin helps you quickly understand long pieces of text by generating concise summaries with extracted key points. It's perfect for:
-
-- Summarizing long articles or documents
-- Extracting key points from conversations
-- Creating quick overviews of complex topics
-
-## Features
-
-- :material-text-box-search: **Smart Summarization**: AI-powered content analysis
-- :material-format-list-bulleted: **Key Points**: Extracted important highlights
-- :material-content-copy: **Easy Copy**: One-click copying of summaries
-- :material-tune: **Adjustable Length**: Control summary detail level
-
----
-
-## Installation
-
-1. Download the plugin file: [`summary.py`](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/summary)
-2. Upload to OpenWebUI: **Admin Panel** → **Settings** → **Functions**
-3. Enable the plugin
-
----
-
-## Usage
-
-1. Get a long response from the AI or paste long text
-2. Click the **Summary** button in the message action bar
-3. View the generated summary with key points
-
----
-
-## Configuration
-
-| Option | Type | Default | Description |
-|--------|------|---------|-------------|
-| `summary_length` | string | `"medium"` | Length of summary (short/medium/long) |
-| `include_key_points` | boolean | `true` | Extract and list key points |
-| `language` | string | `"auto"` | Output language |
-
----
-
-## Example Output
-
-```markdown
-## Summary
-
-This document discusses the implementation of a new feature 
-for the application, focusing on user experience improvements 
-and performance optimizations.
-
-### Key Points
-
-- ✅ New user interface design improves accessibility
-- ✅ Backend optimizations reduce load times by 40%
-- ✅ Mobile responsiveness enhanced
-- ✅ Integration with third-party services simplified
-```
-
----
-
-## Requirements
-
-!!! note "Prerequisites"
-    - OpenWebUI v0.3.0 or later
-    - Uses the active LLM model for summarization
-
----
-
-## Source Code
-
-[:fontawesome-brands-github: View on GitHub](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/summary){ .md-button }

docs/plugins/actions/summary.zh.md

@@ -1,82 +0,0 @@
-# Summary（摘要）
-
-<span class="category-badge action">Action</span>
-<span class="version-badge">v0.1.0</span>
-
-为长文本生成简洁摘要，并提取关键要点。
-
----
-
-## 概览
-
-Summary 插件可以快速理解长文本，生成精炼摘要并列出关键点，适合：
-
-- 总结长文章或文档
-- 从对话中提炼要点
-- 为复杂主题制作快速概览
-
-## 功能特性
-
-- :material-text-box-search: **智能摘要**：AI 驱动的内容分析
-- :material-format-list-bulleted: **关键点**：提取重要信息
-- :material-content-copy: **便捷复制**：一键复制摘要
-- :material-tune: **长度可调**：可选择摘要详略程度
-
----
-
-## 安装
-
-1. 下载插件文件：[`summary.py`](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/summary)
-2. 上传到 OpenWebUI：**Admin Panel** → **Settings** → **Functions**
-3. 启用插件
-
----
-
-## 使用方法
-
-1. 获取一段较长的 AI 回复或粘贴长文本
-2. 点击消息操作栏的 **Summary** 按钮
-3. 查看生成的摘要与关键点
-
----
-
-## 配置项
-
-| 选项 | 类型 | 默认值 | 说明 |
-|--------|------|---------|-------------|
-| `summary_length` | string | `"medium"` | 摘要长度（short/medium/long） |
-| `include_key_points` | boolean | `true` | 是否提取并列出关键点 |
-| `language` | string | `"auto"` | 输出语言 |
-
----
-
-## 输出示例
-
-```markdown
-## Summary
-
-This document discusses the implementation of a new feature 
-for the application, focusing on user experience improvements 
-and performance optimizations.
-
-### Key Points
-
-- ✅ New user interface design improves accessibility
-- ✅ Backend optimizations reduce load times by 40%
-- ✅ Mobile responsiveness enhanced
-- ✅ Integration with third-party services simplified
-```
-
----
-
-## 运行要求
-
-!!! note "前置条件"
-    - OpenWebUI v0.3.0 及以上
-    - 使用当前会话的 LLM 模型进行摘要
-
----
-
-## 源码
-
-[:fontawesome-brands-github: 在 GitHub 查看](https://github.com/Fu-Jie/awesome-openwebui/tree/main/plugins/actions/summary){ .md-button }

docs/plugins/filters/async-context-compression.md

@@ -1,7 +1,7 @@
 # Async Context Compression
 
 <span class="category-badge filter">Filter</span>
-<span class="version-badge">v1.1.0</span>
+<span class="version-badge">v1.1.3</span>
 
 Reduces token consumption in long conversations through intelligent summarization while maintaining conversational coherence.
 
@@ -29,6 +29,11 @@ This is especially useful for:
 - :material-clock-fast: **Async Processing**: Non-blocking background compression
 - :material-memory: **Context Preservation**: Keeps important information
 - :material-currency-usd-off: **Cost Reduction**: Minimize token usage
+- :material-console: **Frontend Debugging**: Debug logs in browser console
+- :material-alert-circle-check: **Enhanced Error Reporting**: Clear error status notifications
+- :material-check-all: **Open WebUI v0.7.x Compatibility**: Dynamic DB session handling
+- :material-account-convert: **Improved Compatibility**: Summary role changed to `assistant`
+- :material-shield-check: **Enhanced Stability**: Resolved race conditions in state management
 
 ---
 

docs/plugins/filters/async-context-compression.zh.md

@@ -1,7 +1,7 @@
 # Async Context Compression（异步上下文压缩）
 
 <span class="category-badge filter">Filter</span>
-<span class="version-badge">v1.1.0</span>
+<span class="version-badge">v1.1.3</span>
 
 通过智能摘要减少长对话的 token 消耗，同时保持对话连贯。
 
@@ -29,6 +29,11 @@ Async Context Compression 过滤器通过以下方式帮助管理长对话的 to
 - :material-clock-fast: **异步处理**：后台非阻塞压缩
 - :material-memory: **保留上下文**：尽量保留重要信息
 - :material-currency-usd-off: **降低成本**：减少 token 使用
+- :material-console: **前端调试**：支持浏览器控制台日志
+- :material-alert-circle-check: **增强错误报告**：清晰的错误状态通知
+- :material-check-all: **Open WebUI v0.7.x 兼容性**：动态数据库会话处理
+- :material-account-convert: **兼容性提升**：摘要角色改为 `assistant`
+- :material-shield-check: **稳定性增强**：解决状态管理竞态条件
 
 ---
 

docs/plugins/filters/index.md

@@ -22,7 +22,7 @@ Filters act as middleware in the message pipeline:
 
     Reduces token consumption in long conversations through intelligent summarization while maintaining coherence.
 
-    **Version:** 1.1.0
+    **Version:** 1.1.3
 
     [:octicons-arrow-right-24: Documentation](async-context-compression.md)
 
@@ -46,6 +46,16 @@ Filters act as middleware in the message pipeline:
 
     [:octicons-arrow-right-24: Documentation](gemini-manifold-companion.md)
 
+-   :material-format-paint:{ .lg .middle } **Markdown Normalizer**
+
+    ---
+
+    Fixes common Markdown formatting issues in LLM outputs, including Mermaid syntax, code blocks, and LaTeX formulas.
+
+    **Version:** 1.0.1
+
+    [:octicons-arrow-right-24: Documentation](markdown_normalizer.md)
+
 </div>
 
 ---

docs/plugins/filters/index.zh.md

@@ -22,7 +22,7 @@ Filter 充当消息管线中的中间件：
 
     通过智能总结减少长对话的 token 消耗，同时保持连贯性。
 
-    **版本：** 1.1.0
+    **版本：** 1.1.3
 
     [:octicons-arrow-right-24: 查看文档](async-context-compression.md)
 
@@ -46,6 +46,16 @@ Filter 充当消息管线中的中间件：
 
     [:octicons-arrow-right-24: 查看文档](gemini-manifold-companion.md)
 
+-   :material-format-paint:{ .lg .middle } **Markdown Normalizer**
+
+    ---
+
+    修复 LLM 输出中常见的 Markdown 格式问题，包括 Mermaid 语法、代码块和 LaTeX 公式。
+
+    **版本：** 1.0.1
+
+    [:octicons-arrow-right-24: 查看文档](markdown_normalizer.zh.md)
+
 </div>
 
 ---

docs/plugins/filters/markdown_normalizer.md

@@ -0,0 +1,46 @@
+# Markdown Normalizer Filter
+
+A production-grade content normalizer filter for Open WebUI that fixes common Markdown formatting issues in LLM outputs. It ensures that code blocks, LaTeX formulas, Mermaid diagrams, and other Markdown elements are rendered correctly.
+
+## Features
+
+*   **Mermaid Syntax Fix**: Automatically fixes common Mermaid syntax errors, such as unquoted node labels (including multi-line labels and citations) and unclosed subgraphs, ensuring diagrams render correctly.
+*   **Frontend Console Debugging**: Supports printing structured debug logs directly to the browser console (F12) for easier troubleshooting.
+*   **Code Block Formatting**: Fixes broken code block prefixes, suffixes, and indentation.
+*   **LaTeX Normalization**: Standardizes LaTeX formula delimiters (`\[` -> `$$`, `\(` -> `$`).
+*   **Thought Tag Normalization**: Unifies thought tags (`<think>`, `<thinking>` -> `<thought>`).
+*   **Escape Character Fix**: Cleans up excessive escape characters (`\\n`, `\\t`).
+*   **List Formatting**: Ensures proper newlines in list items.
+*   **Heading Fix**: Adds missing spaces in headings (`#Heading` -> `# Heading`).
+*   **Table Fix**: Adds missing closing pipes in tables.
+*   **XML Cleanup**: Removes leftover XML artifacts.
+
+## Usage
+
+1.  Install the plugin in Open WebUI.
+2.  Enable the filter globally or for specific models.
+3.  Configure the enabled fixes in the **Valves** settings.
+4.  (Optional) **Show Debug Log** is enabled by default in Valves. This prints structured logs to the browser console (F12).
+    > [!WARNING]
+    > As this is an initial version, some "negative fixes" might occur (e.g., breaking valid Markdown). If you encounter issues, please check the console logs, copy the "Original" vs "Normalized" content, and submit an issue.
+
+## Configuration (Valves)
+
+*   `priority`: Filter priority (default: 50).
+*   `enable_escape_fix`: Fix excessive escape characters.
+*   `enable_thought_tag_fix`: Normalize thought tags.
+*   `enable_code_block_fix`: Fix code block formatting.
+*   `enable_latex_fix`: Normalize LaTeX formulas.
+*   `enable_list_fix`: Fix list item newlines (Experimental).
+*   `enable_unclosed_block_fix`: Auto-close unclosed code blocks.
+*   `enable_fullwidth_symbol_fix`: Fix full-width symbols in code blocks.
+*   `enable_mermaid_fix`: Fix Mermaid syntax errors.
+*   `enable_heading_fix`: Fix missing space in headings.
+*   `enable_table_fix`: Fix missing closing pipe in tables.
+*   `enable_xml_tag_cleanup`: Cleanup leftover XML tags.
+*   `show_status`: Show status notification when fixes are applied.
+*   `show_debug_log`: Print debug logs to browser console.
+
+## License
+
+MIT

docs/plugins/filters/markdown_normalizer.zh.md

@@ -0,0 +1,46 @@
+# Markdown 格式化过滤器 (Markdown Normalizer)
+
+这是一个用于 Open WebUI 的生产级内容格式化过滤器，旨在修复 LLM 输出中常见的 Markdown 格式问题。它能确保代码块、LaTeX 公式、Mermaid 图表和其他 Markdown 元素被正确渲染。
+
+## 功能特性
+
+*   **Mermaid 语法修复**: 自动修复常见的 Mermaid 语法错误，如未加引号的节点标签（支持多行标签和引用标记）和未闭合的子图 (Subgraph)，确保图表能正确渲染。
+*   **前端控制台调试**: 支持将结构化的调试日志直接打印到浏览器控制台 (F12)，方便排查问题。
+*   **代码块格式化**: 修复破损的代码块前缀、后缀和缩进问题。
+*   **LaTeX 规范化**: 标准化 LaTeX 公式定界符 (`\[` -> `$$`, `\(` -> `$`)。
+*   **思维标签规范化**: 统一思维链标签 (`<think>`, `<thinking>` -> `<thought>`)。
+*   **转义字符修复**: 清理过度的转义字符 (`\\n`, `\\t`)。
+*   **列表格式化**: 确保列表项有正确的换行。
+*   **标题修复**: 修复标题中缺失的空格 (`#标题` -> `# 标题`)。
+*   **表格修复**: 修复表格中缺失的闭合管道符。
+*   **XML 清理**: 移除残留的 XML 标签。
+
+## 使用方法
+
+1.  在 Open WebUI 中安装此插件。
+2.  全局启用或为特定模型启用此过滤器。
+3.  在 **Valves** 设置中配置需要启用的修复项。
+4.  (可选) **显示调试日志 (Show Debug Log)** 在 Valves 中默认开启。这会将结构化的日志打印到浏览器控制台 (F12)。
+    > [!WARNING]
+    > 由于这是初版，可能会出现“负向修复”的情况（例如破坏了原本正确的格式）。如果您遇到问题，请务必查看控制台日志，复制“原始 (Original)”与“规范化 (Normalized)”的内容对比，并提交 Issue 反馈。
+
+## 配置项 (Valves)
+
+*   `priority`: 过滤器优先级 (默认: 50)。
+*   `enable_escape_fix`: 修复过度的转义字符。
+*   `enable_thought_tag_fix`: 规范化思维标签。
+*   `enable_code_block_fix`: 修复代码块格式。
+*   `enable_latex_fix`: 规范化 LaTeX 公式。
+*   `enable_list_fix`: 修复列表项换行 (实验性)。
+*   `enable_unclosed_block_fix`: 自动闭合未闭合的代码块。
+*   `enable_fullwidth_symbol_fix`: 修复代码块中的全角符号。
+*   `enable_mermaid_fix`: 修复 Mermaid 语法错误。
+*   `enable_heading_fix`: 修复标题中缺失的空格。
+*   `enable_table_fix`: 修复表格中缺失的闭合管道符。
+*   `enable_xml_tag_cleanup`: 清理残留的 XML 标签。
+*   `show_status`: 应用修复时显示状态通知。
+*   `show_debug_log`: 在浏览器控制台打印调试日志。
+
+## 许可证
+
+MIT

docs/plugins/index.md

@@ -53,7 +53,6 @@ OpenWebUI supports four types of plugins, each serving a different purpose:
 | [Knowledge Card](actions/knowledge-card.md) | Action | Create beautiful learning flashcards | 0.2.0 |
 | [Export to Excel](actions/export-to-excel.md) | Action | Export chat history to Excel files | 1.0.0 |
 | [Export to Word](actions/export-to-word.md) | Action | Export chat content to Word (.docx) with formatting | 0.1.0 |
-| [Summary](actions/summary.md) | Action | Text summarization tool | 1.0.0 |
 | [Async Context Compression](filters/async-context-compression.md) | Filter | Intelligent context compression | 1.0.0 |
 | [Context Enhancement](filters/context-enhancement.md) | Filter | Enhance chat context | 1.0.0 |
 | [Gemini Manifold Companion](filters/gemini-manifold-companion.md) | Filter | Companion for Gemini Manifold | 1.0.0 |

docs/plugins/index.zh.md

@@ -53,7 +53,6 @@ OpenWebUI 支持四种类型的插件，每种都有不同的用途：
 | [Knowledge Card（知识卡片）](actions/knowledge-card.md) | Action | 生成精美学习卡片 | 0.2.0 |
 | [Export to Excel（导出到 Excel）](actions/export-to-excel.md) | Action | 导出聊天记录为 Excel | 1.0.0 |
 | [Export to Word（导出为 Word）](actions/export-to-word.md) | Action | 将聊天内容导出为 Word (.docx) 并保留格式 | 0.1.0 |
-| [Summary（摘要）](actions/summary.md) | Action | 文本摘要工具 | 1.0.0 |
 | [Async Context Compression（异步上下文压缩）](filters/async-context-compression.md) | Filter | 智能上下文压缩 | 1.0.0 |
 | [Context Enhancement（上下文增强）](filters/context-enhancement.md) | Filter | 提升对话上下文 | 1.0.0 |
 | [Gemini Manifold Companion](filters/gemini-manifold-companion.md) | Filter | Gemini Manifold 伴侣 | 1.0.0 |

mkdocs.yml

@@ -187,7 +187,6 @@ nav:
       - Knowledge Card: plugins/actions/knowledge-card.md
       - Export to Excel: plugins/actions/export-to-excel.md
       - Export to Word: plugins/actions/export-to-word.md
-      - Summary: plugins/actions/summary.md
     - Filters:
       - plugins/filters/index.md
       - Async Context Compression: plugins/filters/async-context-compression.md

plugins/actions/ACTION_PLUGIN_TEMPLATE.py

@@ -315,7 +315,7 @@ class Action:
                     if role == "user"
                     else "Assistant" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} Message {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             return body  # Or handle error

plugins/actions/ACTION_PLUGIN_TEMPLATE_CN.py

@@ -326,7 +326,7 @@ class Action:
                     if role == "user"
                     else "助手" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} 消息 {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             return body  # 或者处理错误

plugins/actions/deep-dive/README.md

@@ -0,0 +1,83 @@
+# 🌊 Deep Dive
+
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.0.0 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
+
+A comprehensive thinking lens that dives deep into any content - from context to logic, insights, and action paths.
+
+## 🔥 What's New in v1.0.0
+
+- ✨ **Thinking Chain Structure**: Moves from surface understanding to deep strategic action.
+- 🔍 **Phase 01: The Context**: Panoramic view of the situation and background.
+- 🧠 **Phase 02: The Logic**: Deconstruction of the underlying reasoning and mental models.
+- 💎 **Phase 03: The Insight**: Extraction of non-obvious value and hidden implications.
+- 🚀 **Phase 04: The Path**: Definition of specific, prioritized strategic directions.
+- 🎨 **Premium UI**: Modern, process-oriented design with a "Thinking Line" timeline.
+- 🌗 **Theme Adaptive**: Automatically adapts to OpenWebUI's light/dark theme.
+
+## ✨ Key Features
+
+- 🌊 **Deep Thinking**: Not just a summary, but a full deconstruction of content.
+- 🧠 **Logical Analysis**: Reveals how arguments are built and identifies hidden assumptions.
+- 💎 **Value Extraction**: Finds the "Aha!" moments and blind spots.
+- 🚀 **Action Oriented**: Translates deep understanding into immediate, actionable steps.
+- 🌍 **Multi-language**: Automatically adapts to the user's preferred language.
+- 🌗 **Theme Support**: Seamlessly switches between light and dark themes based on OpenWebUI settings.
+
+## 🚀 How to Use
+
+1. **Input Content**: Provide any text, article, or meeting notes in the chat.
+2. **Trigger Deep Dive**: Click the **Deep Dive** action button.
+3. **Explore the Chain**: Follow the visual timeline from Context to Path.
+
+## ⚙️ Configuration (Valves)
+
+| Parameter | Default | Description |
+| :--- | :--- | :--- |
+| **Show Status (SHOW_STATUS)** | `True` | Whether to show status updates during the thinking process. |
+| **Model ID (MODEL_ID)** | `Empty` | LLM model for analysis. Empty = use current model. |
+| **Min Text Length (MIN_TEXT_LENGTH)** | `200` | Minimum characters required for a meaningful deep dive. |
+| **Clear Previous HTML (CLEAR_PREVIOUS_HTML)** | `True` | Whether to clear previous plugin results. |
+| **Message Count (MESSAGE_COUNT)** | `1` | Number of recent messages to analyze. |
+
+## 🌗 Theme Support
+
+The plugin automatically detects and adapts to OpenWebUI's theme settings:
+
+- **Detection Priority**:
+  1. Parent document `<meta name="theme-color">` tag
+  2. Parent document `html/body` class or `data-theme` attribute
+  3. System preference via `prefers-color-scheme: dark`
+
+- **Requirements**: For best results, enable **iframe Sandbox Allow Same Origin** in OpenWebUI: 
+  - Go to **Settings** → **Interface** → **Artifacts** → Check **iframe Sandbox Allow Same Origin**
+
+## 🎨 Visual Preview
+
+The plugin generates a structured thinking timeline:
+
+```
+┌─────────────────────────────────────┐
+│  🌊 Deep Dive Analysis              │
+│  👤 User  📅 Date  📊 Word count    │
+├─────────────────────────────────────┤
+│  🔍 Phase 01: The Context           │
+│  [High-level panoramic view]        │
+│                                     │
+│  🧠 Phase 02: The Logic             │
+│  • Reasoning structure...           │
+│  • Hidden assumptions...            │
+│                                     │
+│  💎 Phase 03: The Insight           │
+│  • Non-obvious value...             │
+│  • Blind spots revealed...          │
+│                                     │
+│  🚀 Phase 04: The Path              │
+│  ▸ Priority Action 1...             │
+│  ▸ Priority Action 2...             │
+└─────────────────────────────────────┘
+```
+
+## 📂 Files
+
+- `deep_dive.py` - English version
+- `deep_dive_cn.py` - Chinese version (精读)

plugins/actions/deep-dive/README_CN.md

@@ -0,0 +1,83 @@
+# 📖 精读
+
+**作者:** [Fu-Jie](https://github.com/Fu-Jie) | **版本:** 1.0.0 | **项目:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
+
+全方位的思维透镜 —— 从背景全景到逻辑脉络，从深度洞察到行动路径。
+
+## 🔥 v1.0.0 更新内容
+
+- ✨ **思维链结构**: 从表面理解一步步深入到战略行动。
+- 🔍 **阶段 01: 全景 (The Context)**: 提供情境与背景的高层级全景视图。
+- 🧠 **阶段 02: 脉络 (The Logic)**: 解构底层推理逻辑与思维模型。
+- 💎 **阶段 03: 洞察 (The Insight)**: 提取非显性价值与隐藏的深层含义。
+- 🚀 **阶段 04: 路径 (The Path)**: 定义具体的、按优先级排列的战略方向。
+- 🎨 **高端 UI**: 现代化的过程导向设计，带有"思维导火索"时间轴。
+- 🌗 **主题自适应**: 自动适配 OpenWebUI 的深色/浅色主题。
+
+## ✨ 核心特性
+
+- 📖 **深度思考**: 不仅仅是摘要，而是对内容的全面解构。
+- 🧠 **逻辑分析**: 揭示论点是如何构建的，识别隐藏的假设。
+- 💎 **价值提取**: 发现"原来如此"的时刻与思维盲点。
+- 🚀 **行动导向**: 将深度理解转化为立即、可执行的步骤。
+- 🌍 **多语言支持**: 自动适配用户的偏好语言。
+- 🌗 **主题支持**: 根据 OpenWebUI 设置自动切换深色/浅色主题。
+
+## 🚀 如何使用
+
+1. **输入内容**: 在聊天中提供任何文本、文章或会议记录。
+2. **触发精读**: 点击 **精读** 操作按钮。
+3. **探索思维链**: 沿着视觉时间轴从"全景"探索到"路径"。
+
+## ⚙️ 配置参数 (Valves)
+
+| 参数 | 默认值 | 描述 |
+| :--- | :--- | :--- |
+| **显示状态 (SHOW_STATUS)** | `True` | 是否在思维过程中显示状态更新。 |
+| **模型 ID (MODEL_ID)** | `空` | 用于分析的 LLM 模型。留空 = 使用当前模型。 |
+| **最小文本长度 (MIN_TEXT_LENGTH)** | `200` | 进行有意义的精读所需的最小字符数。 |
+| **清除旧 HTML (CLEAR_PREVIOUS_HTML)** | `True` | 是否清除之前的插件结果。 |
+| **消息数量 (MESSAGE_COUNT)** | `1` | 要分析的最近消息数量。 |
+
+## 🌗 主题支持
+
+插件会自动检测并适配 OpenWebUI 的主题设置：
+
+- **检测优先级**:
+  1. 父文档 `<meta name="theme-color">` 标签
+  2. 父文档 `html/body` 的 class 或 `data-theme` 属性
+  3. 系统偏好 `prefers-color-scheme: dark`
+
+- **环境要求**: 为获得最佳效果，请在 OpenWebUI 中启用 **iframe Sandbox Allow Same Origin**：
+  - 进入 **设置** → **界面** → **Artifacts** → 勾选 **iframe Sandbox Allow Same Origin**
+
+## 🎨 视觉预览
+
+插件生成结构化的思维时间轴：
+
+```
+┌─────────────────────────────────────┐
+│  📖 精读分析报告                     │
+│  👤 用户  📅 日期  📊 字数           │
+├─────────────────────────────────────┤
+│  🔍 阶段 01: 全景 (The Context)      │
+│  [高层级全景视图内容]                 │
+│                                     │
+│  🧠 阶段 02: 脉络 (The Logic)        │
+│  • 推理结构分析...                   │
+│  • 隐藏假设识别...                   │
+│                                     │
+│  💎 阶段 03: 洞察 (The Insight)      │
+│  • 非显性价值提取...                 │
+│  • 思维盲点揭示...                   │
+│                                     │
+│  🚀 阶段 04: 路径 (The Path)         │
+│  ▸ 优先级行动 1...                   │
+│  ▸ 优先级行动 2...                   │
+└─────────────────────────────────────┘
+```
+
+## 📂 文件说明
+
+- `deep_dive.py` - 英文版 (Deep Dive)
+- `deep_dive_cn.py` - 中文版 (精读)

plugins/actions/deep-dive/deep_dive.py

@@ -0,0 +1,884 @@
+"""
+title: Deep Dive
+author: Fu-Jie
+author_url: https://github.com/Fu-Jie
+funding_url: https://github.com/Fu-Jie/awesome-openwebui
+version: 1.0.0
+icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxwYXRoIGQ9Ik0xMiA3djE0Ii8+PHBhdGggZD0iTTMgMThhMSAxIDAgMCAxLTEtMVY0YTEgMSAwIDAgMSAxLTFoNWE0IDQgMCAwIDEgNCA0IDQgNCAwIDAgMSA0LTRoNWExIDEgMCAwIDEgMSAxdjEzYTEgMSAwIDAgMS0xIDFoLTZhMyAzIDAgMCAwLTMgMyAzIDMgMCAwIDAtMy0zeiIvPjxwYXRoIGQ9Ik02IDEyaDIiLz48cGF0aCBkPSJNMTYgMTJoMiIvPjwvc3ZnPg==
+requirements: markdown
+description: A comprehensive thinking lens that dives deep into any content - from context to logic, insights, and action paths.
+"""
+
+# Standard library imports
+import re
+import logging
+from typing import Optional, Dict, Any, Callable, Awaitable
+from datetime import datetime
+
+# Third-party imports
+from pydantic import BaseModel, Field
+from fastapi import Request
+import markdown
+
+# OpenWebUI imports
+from open_webui.utils.chat import generate_chat_completion
+from open_webui.models.users import Users
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+)
+logger = logging.getLogger(__name__)
+
+# =================================================================
+# HTML Template - Process-Oriented Design with Theme Support
+# =================================================================
+HTML_WRAPPER_TEMPLATE = """
+<!-- OPENWEBUI_PLUGIN_OUTPUT -->
+<!DOCTYPE html>
+<html lang="{user_language}">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <style>
+        :root {
+            --dd-bg-primary: #ffffff;
+            --dd-bg-secondary: #f8fafc;
+            --dd-bg-tertiary: #f1f5f9;
+            --dd-text-primary: #0f172a;
+            --dd-text-secondary: #334155;
+            --dd-text-dim: #64748b;
+            --dd-border: #e2e8f0;
+            --dd-accent: #3b82f6;
+            --dd-accent-soft: #eff6ff;
+            --dd-header-gradient: linear-gradient(135deg, #1e293b 0%, #0f172a 100%);
+            --dd-shadow: 0 10px 40px rgba(0,0,0,0.06);
+            --dd-code-bg: #f1f5f9;
+        }
+        .theme-dark {
+            --dd-bg-primary: #1e293b;
+            --dd-bg-secondary: #0f172a;
+            --dd-bg-tertiary: #334155;
+            --dd-text-primary: #f1f5f9;
+            --dd-text-secondary: #e2e8f0;
+            --dd-text-dim: #94a3b8;
+            --dd-border: #475569;
+            --dd-accent: #60a5fa;
+            --dd-accent-soft: rgba(59, 130, 246, 0.15);
+            --dd-header-gradient: linear-gradient(135deg, #0f172a 0%, #1e1e2e 100%);
+            --dd-shadow: 0 10px 40px rgba(0,0,0,0.3);
+            --dd-code-bg: #334155;
+        }
+        body { 
+            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; 
+            margin: 0; 
+            padding: 10px; 
+            background-color: transparent; 
+        }
+        #main-container { 
+            display: flex; 
+            flex-direction: column;
+            gap: 24px; 
+            width: 100%;
+            max-width: 900px;
+            margin: 0 auto;
+        }
+        .plugin-item { 
+            background: var(--dd-bg-primary); 
+            border-radius: 24px; 
+            box-shadow: var(--dd-shadow); 
+            overflow: hidden; 
+            border: 1px solid var(--dd-border); 
+        }
+        /* STYLES_INSERTION_POINT */
+    </style>
+</head>
+<body>
+    <div id="main-container">
+        <!-- CONTENT_INSERTION_POINT -->
+    </div>
+    <!-- SCRIPTS_INSERTION_POINT -->
+    <script>
+    (function() {
+        const parseColorLuma = (colorStr) => {
+            if (!colorStr) return null;
+            let m = colorStr.match(/^#?([0-9a-f]{6})$/i);
+            if (m) {
+                const hex = m[1];
+                const r = parseInt(hex.slice(0, 2), 16);
+                const g = parseInt(hex.slice(2, 4), 16);
+                const b = parseInt(hex.slice(4, 6), 16);
+                return (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+            }
+            m = colorStr.match(/rgba?\\s*\\(\\s*(\\d+)\\s*,\\s*(\\d+)\\s*,\\s*(\\d+)/i);
+            if (m) {
+                const r = parseInt(m[1], 10);
+                const g = parseInt(m[2], 10);
+                const b = parseInt(m[3], 10);
+                return (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+            }
+            return null;
+        };
+        const getThemeFromMeta = (doc) => {
+            const metas = Array.from((doc || document).querySelectorAll('meta[name="theme-color"]'));
+            if (!metas.length) return null;
+            const color = metas[metas.length - 1].content.trim();
+            const luma = parseColorLuma(color);
+            if (luma === null) return null;
+            return luma < 0.5 ? 'dark' : 'light';
+        };
+        const getParentDocumentSafe = () => {
+            try {
+                if (!window.parent || window.parent === window) return null;
+                const pDoc = window.parent.document;
+                void pDoc.title;
+                return pDoc;
+            } catch (err) { return null; }
+        };
+        const getThemeFromParentClass = () => {
+            try {
+                if (!window.parent || window.parent === window) return null;
+                const pDoc = window.parent.document;
+                const html = pDoc.documentElement;
+                const body = pDoc.body;
+                const htmlClass = html ? html.className : '';
+                const bodyClass = body ? body.className : '';
+                const htmlDataTheme = html ? html.getAttribute('data-theme') : '';
+                if (htmlDataTheme === 'dark' || bodyClass.includes('dark') || htmlClass.includes('dark')) return 'dark';
+                if (htmlDataTheme === 'light' || bodyClass.includes('light') || htmlClass.includes('light')) return 'light';
+                return null;
+            } catch (err) { return null; }
+        };
+        const setTheme = () => {
+            const parentDoc = getParentDocumentSafe();
+            const metaTheme = parentDoc ? getThemeFromMeta(parentDoc) : null;
+            const parentClassTheme = getThemeFromParentClass();
+            const prefersDark = window.matchMedia && window.matchMedia('(prefers-color-scheme: dark)').matches;
+            const chosen = metaTheme || parentClassTheme || (prefersDark ? 'dark' : 'light');
+            document.documentElement.classList.toggle('theme-dark', chosen === 'dark');
+        };
+        setTheme();
+        if (window.matchMedia) {
+            window.matchMedia('(prefers-color-scheme: dark)').addEventListener('change', setTheme);
+        }
+    })();
+    </script>
+</body>
+</html>
+"""
+
+# =================================================================
+# LLM Prompts - Deep Dive Thinking Chain
+# =================================================================
+
+SYSTEM_PROMPT = """
+You are a Deep Dive Analyst. Your goal is to guide the user through a comprehensive thinking process, moving from surface understanding to deep strategic action.
+
+## Thinking Structure (STRICT)
+
+You MUST analyze the input across these four specific dimensions:
+
+### 1. 🔍 The Context (What?)
+Provide a high-level panoramic view. What is this content about? What is the core situation, background, or problem being addressed? (2-3 paragraphs)
+
+### 2. 🧠 The Logic (Why?)
+Deconstruct the underlying structure. How is the argument built? What is the reasoning, the hidden assumptions, or the mental models at play? (Bullet points)
+
+### 3. 💎 The Insight (So What?)
+Extract the non-obvious value. What are the "Aha!" moments? What are the implications, the blind spots, or the unique perspectives revealed? (Bullet points)
+
+### 4. 🚀 The Path (Now What?)
+Define the strategic direction. What are the specific, prioritized next steps? How can this knowledge be applied immediately? (Actionable steps)
+
+## Rules
+- Output in the user's specified language.
+- Maintain a professional, analytical, yet inspiring tone.
+- Focus on the *process* of understanding, not just the result.
+- No greetings or meta-commentary.
+"""
+
+USER_PROMPT = """
+Initiate a Deep Dive into the following content:
+
+**User Context:**
+- User: {user_name}
+- Time: {current_date_time_str}
+- Language: {user_language}
+
+**Content to Analyze:**
+```
+{long_text_content}
+```
+
+Please execute the full thinking chain: Context → Logic → Insight → Path.
+"""
+
+# =================================================================
+# Premium CSS Design - Deep Dive Theme
+# =================================================================
+
+CSS_TEMPLATE = """
+        .deep-dive {
+            font-family: 'Inter', -apple-system, system-ui, sans-serif;
+            color: var(--dd-text-secondary);
+        }
+
+        .dd-header {
+            background: var(--dd-header-gradient);
+            padding: 40px 32px;
+            color: white;
+            position: relative;
+        }
+
+        .dd-header-badge {
+            display: inline-block;
+            padding: 4px 12px;
+            background: rgba(255,255,255,0.1);
+            border: 1px solid rgba(255,255,255,0.2);
+            border-radius: 100px;
+            font-size: 0.75rem;
+            font-weight: 600;
+            letter-spacing: 0.05em;
+            text-transform: uppercase;
+            margin-bottom: 16px;
+        }
+
+        .dd-title {
+            font-size: 2rem;
+            font-weight: 800;
+            margin: 0 0 12px 0;
+            letter-spacing: -0.02em;
+        }
+
+        .dd-meta {
+            display: flex;
+            gap: 20px;
+            font-size: 0.85rem;
+            opacity: 0.7;
+        }
+
+        .dd-body {
+            padding: 32px;
+            display: flex;
+            flex-direction: column;
+            gap: 40px;
+            position: relative;
+            background: var(--dd-bg-primary);
+        }
+
+        /* The Thinking Line */
+        .dd-body::before {
+            content: '';
+            position: absolute;
+            left: 52px;
+            top: 40px;
+            bottom: 40px;
+            width: 2px;
+            background: var(--dd-border);
+            z-index: 0;
+        }
+
+        .dd-step {
+            position: relative;
+            z-index: 1;
+            display: flex;
+            gap: 24px;
+        }
+
+        .dd-step-icon {
+            flex-shrink: 0;
+            width: 40px;
+            height: 40px;
+            background: var(--dd-bg-primary);
+            border: 2px solid var(--dd-border);
+            border-radius: 12px;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            font-size: 1.25rem;
+            box-shadow: 0 4px 12px rgba(0,0,0,0.03);
+            transition: all 0.3s ease;
+        }
+
+        .dd-step:hover .dd-step-icon {
+            border-color: var(--dd-accent);
+            transform: scale(1.1);
+        }
+
+        .dd-step-content {
+            flex: 1;
+        }
+
+        .dd-step-label {
+            font-size: 0.75rem;
+            font-weight: 700;
+            color: var(--dd-accent);
+            text-transform: uppercase;
+            letter-spacing: 0.1em;
+            margin-bottom: 4px;
+        }
+
+        .dd-step-title {
+            font-size: 1.25rem;
+            font-weight: 700;
+            color: var(--dd-text-primary);
+            margin: 0 0 16px 0;
+        }
+
+        .dd-text {
+            line-height: 1.7;
+            font-size: 1rem;
+        }
+
+        .dd-text p { margin-bottom: 16px; }
+        .dd-text p:last-child { margin-bottom: 0; }
+
+        .dd-list {
+            list-style: none;
+            padding: 0;
+            margin: 0;
+            display: grid;
+            gap: 12px;
+        }
+
+        .dd-list-item {
+            background: var(--dd-bg-secondary);
+            padding: 16px 20px;
+            border-radius: 12px;
+            border-left: 4px solid var(--dd-border);
+            transition: all 0.2s ease;
+        }
+
+        .dd-list-item:hover {
+            background: var(--dd-bg-tertiary);
+            border-left-color: var(--dd-accent);
+            transform: translateX(4px);
+        }
+
+        .dd-list-item strong {
+            color: var(--dd-text-primary);
+            display: block;
+            margin-bottom: 4px;
+        }
+
+        .dd-path-item {
+            background: var(--dd-accent-soft);
+            border-left-color: var(--dd-accent);
+        }
+
+        .dd-footer {
+            padding: 24px 32px;
+            background: var(--dd-bg-secondary);
+            border-top: 1px solid var(--dd-border);
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            font-size: 0.8rem;
+            color: var(--dd-text-dim);
+        }
+
+        .dd-tag {
+            padding: 2px 8px;
+            background: var(--dd-bg-tertiary);
+            border-radius: 4px;
+            font-weight: 600;
+        }
+
+        .dd-text code,
+        .dd-list-item code {
+            background: var(--dd-code-bg);
+            color: var(--dd-text-primary);
+            padding: 2px 6px;
+            border-radius: 4px;
+            font-family: 'SF Mono', 'Consolas', 'Monaco', monospace;
+            font-size: 0.85em;
+        }
+
+        .dd-list-item em {
+            font-style: italic;
+            color: var(--dd-text-dim);
+        }
+"""
+
+CONTENT_TEMPLATE = """
+        <div class="deep-dive">
+            <div class="dd-header">
+                <div class="dd-header-badge">Thinking Process</div>
+                <h1 class="dd-title">Deep Dive Analysis</h1>
+                <div class="dd-meta">
+                    <span>👤 {user_name}</span>
+                    <span>📅 {current_date_time_str}</span>
+                    <span>📊 {word_count} words</span>
+                </div>
+            </div>
+            <div class="dd-body">
+                <!-- Step 1: Context -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🔍</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 01</div>
+                        <h2 class="dd-step-title">The Context</h2>
+                        <div class="dd-text">{context_html}</div>
+                    </div>
+                </div>
+
+                <!-- Step 2: Logic -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🧠</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 02</div>
+                        <h2 class="dd-step-title">The Logic</h2>
+                        <div class="dd-text">{logic_html}</div>
+                    </div>
+                </div>
+
+                <!-- Step 3: Insight -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">💎</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 03</div>
+                        <h2 class="dd-step-title">The Insight</h2>
+                        <div class="dd-text">{insight_html}</div>
+                    </div>
+                </div>
+
+                <!-- Step 4: Path -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🚀</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 04</div>
+                        <h2 class="dd-step-title">The Path</h2>
+                        <div class="dd-text">{path_html}</div>
+                    </div>
+                </div>
+            </div>
+            <div class="dd-footer">
+                <span>Deep Dive Engine v1.0</span>
+                <span><span class="dd-tag">AI-Powered</span></span>
+            </div>
+        </div>
+"""
+
+
+class Action:
+    class Valves(BaseModel):
+        SHOW_STATUS: bool = Field(
+            default=True,
+            description="Whether to show operation status updates.",
+        )
+        MODEL_ID: str = Field(
+            default="",
+            description="LLM Model ID for analysis. Empty = use current model.",
+        )
+        MIN_TEXT_LENGTH: int = Field(
+            default=200,
+            description="Minimum text length for deep dive (chars).",
+        )
+        CLEAR_PREVIOUS_HTML: bool = Field(
+            default=True,
+            description="Whether to clear previous plugin results.",
+        )
+        MESSAGE_COUNT: int = Field(
+            default=1,
+            description="Number of recent messages to analyze.",
+        )
+
+    def __init__(self):
+        self.valves = self.Valves()
+
+    def _get_user_context(self, __user__: Optional[Dict[str, Any]]) -> Dict[str, str]:
+        """Safely extracts user context information."""
+        if isinstance(__user__, (list, tuple)):
+            user_data = __user__[0] if __user__ else {}
+        elif isinstance(__user__, dict):
+            user_data = __user__
+        else:
+            user_data = {}
+
+        return {
+            "user_id": user_data.get("id", "unknown_user"),
+            "user_name": user_data.get("name", "User"),
+            "user_language": user_data.get("language", "en-US"),
+        }
+
+    def _process_llm_output(self, llm_output: str) -> Dict[str, str]:
+        """Parse LLM output and convert to styled HTML."""
+        # Extract sections using flexible regex
+        context_match = re.search(
+            r"###\s*1\.\s*🔍?\s*The Context\s*\((.*?)\)\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        logic_match = re.search(
+            r"###\s*2\.\s*🧠?\s*The Logic\s*\((.*?)\)\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        insight_match = re.search(
+            r"###\s*3\.\s*💎?\s*The Insight\s*\((.*?)\)\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        path_match = re.search(
+            r"###\s*4\.\s*🚀?\s*The Path\s*\((.*?)\)\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+
+        # Fallback if numbering is different
+        if not context_match:
+            context_match = re.search(
+                r"###\s*🔍?\s*The Context.*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not logic_match:
+            logic_match = re.search(
+                r"###\s*🧠?\s*The Logic.*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not insight_match:
+            insight_match = re.search(
+                r"###\s*💎?\s*The Insight.*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not path_match:
+            path_match = re.search(
+                r"###\s*🚀?\s*The Path.*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+
+        context_md = (
+            context_match.group(1 if context_match.lastindex == 1 else 2).strip()
+            if context_match
+            else ""
+        )
+        logic_md = (
+            logic_match.group(1 if logic_match.lastindex == 1 else 2).strip()
+            if logic_match
+            else ""
+        )
+        insight_md = (
+            insight_match.group(1 if insight_match.lastindex == 1 else 2).strip()
+            if insight_match
+            else ""
+        )
+        path_md = (
+            path_match.group(1 if path_match.lastindex == 1 else 2).strip()
+            if path_match
+            else ""
+        )
+
+        if not any([context_md, logic_md, insight_md, path_md]):
+            context_md = llm_output.strip()
+            logger.warning("LLM output did not follow format. Using as context.")
+
+        md_extensions = ["nl2br"]
+
+        context_html = (
+            markdown.markdown(context_md, extensions=md_extensions)
+            if context_md
+            else '<p class="dd-no-content">No context extracted.</p>'
+        )
+        logic_html = (
+            self._process_list_items(logic_md, "logic")
+            if logic_md
+            else '<p class="dd-no-content">No logic deconstructed.</p>'
+        )
+        insight_html = (
+            self._process_list_items(insight_md, "insight")
+            if insight_md
+            else '<p class="dd-no-content">No insights found.</p>'
+        )
+        path_html = (
+            self._process_list_items(path_md, "path")
+            if path_md
+            else '<p class="dd-no-content">No path defined.</p>'
+        )
+
+        return {
+            "context_html": context_html,
+            "logic_html": logic_html,
+            "insight_html": insight_html,
+            "path_html": path_html,
+        }
+
+    def _process_list_items(self, md_content: str, section_type: str) -> str:
+        """Convert markdown list to styled HTML cards with full markdown support."""
+        lines = md_content.strip().split("\n")
+        items = []
+        current_paragraph = []
+
+        for line in lines:
+            line = line.strip()
+
+            # Check for list item (bullet or numbered)
+            bullet_match = re.match(r"^[-*]\s+(.+)$", line)
+            numbered_match = re.match(r"^\d+\.\s+(.+)$", line)
+
+            if bullet_match or numbered_match:
+                # Flush any accumulated paragraph
+                if current_paragraph:
+                    para_text = " ".join(current_paragraph)
+                    para_html = self._convert_inline_markdown(para_text)
+                    items.append(f"<p>{para_html}</p>")
+                    current_paragraph = []
+
+                # Extract the list item content
+                text = (
+                    bullet_match.group(1) if bullet_match else numbered_match.group(1)
+                )
+
+                # Handle bold title pattern: **Title:** Description or **Title**: Description
+                title_match = re.match(r"\*\*(.+?)\*\*[:\s]*(.*)$", text)
+                if title_match:
+                    title = self._convert_inline_markdown(title_match.group(1))
+                    desc = self._convert_inline_markdown(title_match.group(2).strip())
+                    path_class = "dd-path-item" if section_type == "path" else ""
+                    item_html = f'<div class="dd-list-item {path_class}"><strong>{title}</strong>{desc}</div>'
+                else:
+                    text_html = self._convert_inline_markdown(text)
+                    path_class = "dd-path-item" if section_type == "path" else ""
+                    item_html = (
+                        f'<div class="dd-list-item {path_class}">{text_html}</div>'
+                    )
+                items.append(item_html)
+            elif line and not line.startswith("#"):
+                # Accumulate paragraph text
+                current_paragraph.append(line)
+            elif not line and current_paragraph:
+                # Empty line ends paragraph
+                para_text = " ".join(current_paragraph)
+                para_html = self._convert_inline_markdown(para_text)
+                items.append(f"<p>{para_html}</p>")
+                current_paragraph = []
+
+        # Flush remaining paragraph
+        if current_paragraph:
+            para_text = " ".join(current_paragraph)
+            para_html = self._convert_inline_markdown(para_text)
+            items.append(f"<p>{para_html}</p>")
+
+        if items:
+            return f'<div class="dd-list">{" ".join(items)}</div>'
+        return f'<p class="dd-no-content">No items found.</p>'
+
+    def _convert_inline_markdown(self, text: str) -> str:
+        """Convert inline markdown (bold, italic, code) to HTML."""
+        # Convert inline code: `code` -> <code>code</code>
+        text = re.sub(r"`([^`]+)`", r"<code>\1</code>", text)
+        # Convert bold: **text** -> <strong>text</strong>
+        text = re.sub(r"\*\*(.+?)\*\*", r"<strong>\1</strong>", text)
+        # Convert italic: *text* -> <em>text</em> (but not inside **)
+        text = re.sub(r"(?<!\*)\*([^*]+)\*(?!\*)", r"<em>\1</em>", text)
+        return text
+
+    async def _emit_status(
+        self,
+        emitter: Optional[Callable[[Any], Awaitable[None]]],
+        description: str,
+        done: bool = False,
+    ):
+        """Emits a status update event."""
+        if self.valves.SHOW_STATUS and emitter:
+            await emitter(
+                {"type": "status", "data": {"description": description, "done": done}}
+            )
+
+    async def _emit_notification(
+        self,
+        emitter: Optional[Callable[[Any], Awaitable[None]]],
+        content: str,
+        ntype: str = "info",
+    ):
+        """Emits a notification event."""
+        if emitter:
+            await emitter(
+                {"type": "notification", "data": {"type": ntype, "content": content}}
+            )
+
+    def _remove_existing_html(self, content: str) -> str:
+        """Removes existing plugin-generated HTML."""
+        pattern = r"```html\s*<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?```"
+        return re.sub(pattern, "", content).strip()
+
+    def _extract_text_content(self, content) -> str:
+        """Extract text from message content."""
+        if isinstance(content, str):
+            return content
+        elif isinstance(content, list):
+            text_parts = []
+            for item in content:
+                if isinstance(item, dict) and item.get("type") == "text":
+                    text_parts.append(item.get("text", ""))
+                elif isinstance(item, str):
+                    text_parts.append(item)
+            return "\n".join(text_parts)
+        return str(content) if content else ""
+
+    def _merge_html(
+        self,
+        existing_html: str,
+        new_content: str,
+        new_styles: str = "",
+        user_language: str = "en-US",
+    ) -> str:
+        """Merges new content into HTML container."""
+        if "<!-- OPENWEBUI_PLUGIN_OUTPUT -->" in existing_html:
+            base_html = re.sub(r"^```html\s*", "", existing_html)
+            base_html = re.sub(r"\s*```$", "", base_html)
+        else:
+            base_html = HTML_WRAPPER_TEMPLATE.replace("{user_language}", user_language)
+
+        wrapped = f'<div class="plugin-item">\n{new_content}\n</div>'
+
+        if new_styles:
+            base_html = base_html.replace(
+                "/* STYLES_INSERTION_POINT */",
+                f"{new_styles}\n/* STYLES_INSERTION_POINT */",
+            )
+
+        base_html = base_html.replace(
+            "<!-- CONTENT_INSERTION_POINT -->",
+            f"{wrapped}\n<!-- CONTENT_INSERTION_POINT -->",
+        )
+
+        return base_html.strip()
+
+    def _build_content_html(self, context: dict) -> str:
+        """Build content HTML."""
+        html = CONTENT_TEMPLATE
+        for key, value in context.items():
+            html = html.replace(f"{{{key}}}", str(value))
+        return html
+
+    async def action(
+        self,
+        body: dict,
+        __user__: Optional[Dict[str, Any]] = None,
+        __event_emitter__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __request__: Optional[Request] = None,
+    ) -> Optional[dict]:
+        logger.info("Action: Deep Dive v1.0.0 started")
+
+        user_ctx = self._get_user_context(__user__)
+        user_id = user_ctx["user_id"]
+        user_name = user_ctx["user_name"]
+        user_language = user_ctx["user_language"]
+
+        now = datetime.now()
+        current_date_time_str = now.strftime("%b %d, %Y %H:%M")
+
+        original_content = ""
+        try:
+            messages = body.get("messages", [])
+            if not messages:
+                raise ValueError("No messages found.")
+
+            message_count = min(self.valves.MESSAGE_COUNT, len(messages))
+            recent_messages = messages[-message_count:]
+
+            aggregated_parts = []
+            for msg in recent_messages:
+                text = self._extract_text_content(msg.get("content"))
+                if text:
+                    aggregated_parts.append(text)
+
+            if not aggregated_parts:
+                raise ValueError("No text content found.")
+
+            original_content = "\n\n---\n\n".join(aggregated_parts)
+            word_count = len(original_content.split())
+
+            if len(original_content) < self.valves.MIN_TEXT_LENGTH:
+                msg = f"Content too brief ({len(original_content)} chars). Deep Dive requires at least {self.valves.MIN_TEXT_LENGTH} chars for meaningful analysis."
+                await self._emit_notification(__event_emitter__, msg, "warning")
+                return {"messages": [{"role": "assistant", "content": f"⚠️ {msg}"}]}
+
+            await self._emit_notification(
+                __event_emitter__, "🌊 Initiating Deep Dive thinking process...", "info"
+            )
+            await self._emit_status(
+                __event_emitter__, "🌊 Deep Dive: Analyzing Context & Logic...", False
+            )
+
+            prompt = USER_PROMPT.format(
+                user_name=user_name,
+                current_date_time_str=current_date_time_str,
+                user_language=user_language,
+                long_text_content=original_content,
+            )
+
+            model = self.valves.MODEL_ID or body.get("model")
+            payload = {
+                "model": model,
+                "messages": [
+                    {"role": "system", "content": SYSTEM_PROMPT},
+                    {"role": "user", "content": prompt},
+                ],
+                "stream": False,
+            }
+
+            user_obj = Users.get_user_by_id(user_id)
+            if not user_obj:
+                raise ValueError(f"User not found: {user_id}")
+
+            response = await generate_chat_completion(__request__, payload, user_obj)
+            llm_output = response["choices"][0]["message"]["content"]
+
+            processed = self._process_llm_output(llm_output)
+
+            context = {
+                "user_name": user_name,
+                "current_date_time_str": current_date_time_str,
+                "word_count": word_count,
+                **processed,
+            }
+
+            content_html = self._build_content_html(context)
+
+            # Handle existing HTML
+            existing = ""
+            match = re.search(
+                r"```html\s*(<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?)```",
+                original_content,
+            )
+            if match:
+                existing = match.group(1)
+
+            if self.valves.CLEAR_PREVIOUS_HTML or not existing:
+                original_content = self._remove_existing_html(original_content)
+                final_html = self._merge_html(
+                    "", content_html, CSS_TEMPLATE, user_language
+                )
+            else:
+                original_content = self._remove_existing_html(original_content)
+                final_html = self._merge_html(
+                    existing, content_html, CSS_TEMPLATE, user_language
+                )
+
+            body["messages"][-1][
+                "content"
+            ] = f"{original_content}\n\n```html\n{final_html}\n```"
+
+            await self._emit_status(__event_emitter__, "🌊 Deep Dive complete!", True)
+            await self._emit_notification(
+                __event_emitter__,
+                f"🌊 Deep Dive complete, {user_name}! Thinking chain generated.",
+                "success",
+            )
+
+        except Exception as e:
+            logger.error(f"Deep Dive Error: {e}", exc_info=True)
+            body["messages"][-1][
+                "content"
+            ] = f"{original_content}\n\n❌ **Error:** {str(e)}"
+            await self._emit_status(__event_emitter__, "Deep Dive failed.", True)
+            await self._emit_notification(
+                __event_emitter__, f"Error: {str(e)}", "error"
+            )
+
+        return body

plugins/actions/deep-dive/deep_dive_cn.py

@@ -0,0 +1,876 @@
+"""
+title: 精读
+author: Fu-Jie
+author_url: https://github.com/Fu-Jie
+funding_url: https://github.com/Fu-Jie/awesome-openwebui
+version: 1.0.0
+icon_url: data:image/svg+xml;base64,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
+requirements: markdown
+description: 全方位的思维透镜 —— 从背景全景到逻辑脉络，从深度洞察到行动路径。
+"""
+
+# Standard library imports
+import re
+import logging
+from typing import Optional, Dict, Any, Callable, Awaitable
+from datetime import datetime
+
+# Third-party imports
+from pydantic import BaseModel, Field
+from fastapi import Request
+import markdown
+
+# OpenWebUI imports
+from open_webui.utils.chat import generate_chat_completion
+from open_webui.models.users import Users
+
+# Logging setup
+logging.basicConfig(
+    level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+)
+logger = logging.getLogger(__name__)
+
+# =================================================================
+# HTML 模板 - 过程导向设计，支持主题自适应
+# =================================================================
+HTML_WRAPPER_TEMPLATE = """
+<!-- OPENWEBUI_PLUGIN_OUTPUT -->
+<!DOCTYPE html>
+<html lang="{user_language}">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <style>
+        :root {
+            --dd-bg-primary: #ffffff;
+            --dd-bg-secondary: #f8fafc;
+            --dd-bg-tertiary: #f1f5f9;
+            --dd-text-primary: #0f172a;
+            --dd-text-secondary: #334155;
+            --dd-text-dim: #64748b;
+            --dd-border: #e2e8f0;
+            --dd-accent: #3b82f6;
+            --dd-accent-soft: #eff6ff;
+            --dd-header-gradient: linear-gradient(135deg, #1e293b 0%, #0f172a 100%);
+            --dd-shadow: 0 10px 40px rgba(0,0,0,0.06);
+            --dd-code-bg: #f1f5f9;
+        }
+        .theme-dark {
+            --dd-bg-primary: #1e293b;
+            --dd-bg-secondary: #0f172a;
+            --dd-bg-tertiary: #334155;
+            --dd-text-primary: #f1f5f9;
+            --dd-text-secondary: #e2e8f0;
+            --dd-text-dim: #94a3b8;
+            --dd-border: #475569;
+            --dd-accent: #60a5fa;
+            --dd-accent-soft: rgba(59, 130, 246, 0.15);
+            --dd-header-gradient: linear-gradient(135deg, #0f172a 0%, #1e1e2e 100%);
+            --dd-shadow: 0 10px 40px rgba(0,0,0,0.3);
+            --dd-code-bg: #334155;
+        }
+        body { 
+            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; 
+            margin: 0; 
+            padding: 10px; 
+            background-color: transparent; 
+        }
+        #main-container { 
+            display: flex; 
+            flex-direction: column;
+            gap: 24px; 
+            width: 100%;
+            max-width: 900px;
+            margin: 0 auto;
+        }
+        .plugin-item { 
+            background: var(--dd-bg-primary); 
+            border-radius: 24px; 
+            box-shadow: var(--dd-shadow); 
+            overflow: hidden; 
+            border: 1px solid var(--dd-border); 
+        }
+        /* STYLES_INSERTION_POINT */
+    </style>
+</head>
+<body>
+    <div id="main-container">
+        <!-- CONTENT_INSERTION_POINT -->
+    </div>
+    <!-- SCRIPTS_INSERTION_POINT -->
+    <script>
+    (function() {
+        const parseColorLuma = (colorStr) => {
+            if (!colorStr) return null;
+            let m = colorStr.match(/^#?([0-9a-f]{6})$/i);
+            if (m) {
+                const hex = m[1];
+                const r = parseInt(hex.slice(0, 2), 16);
+                const g = parseInt(hex.slice(2, 4), 16);
+                const b = parseInt(hex.slice(4, 6), 16);
+                return (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+            }
+            m = colorStr.match(/rgba?\\s*\\(\\s*(\\d+)\\s*,\\s*(\\d+)\\s*,\\s*(\\d+)/i);
+            if (m) {
+                const r = parseInt(m[1], 10);
+                const g = parseInt(m[2], 10);
+                const b = parseInt(m[3], 10);
+                return (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+            }
+            return null;
+        };
+        const getThemeFromMeta = (doc) => {
+            const metas = Array.from((doc || document).querySelectorAll('meta[name="theme-color"]'));
+            if (!metas.length) return null;
+            const color = metas[metas.length - 1].content.trim();
+            const luma = parseColorLuma(color);
+            if (luma === null) return null;
+            return luma < 0.5 ? 'dark' : 'light';
+        };
+        const getParentDocumentSafe = () => {
+            try {
+                if (!window.parent || window.parent === window) return null;
+                const pDoc = window.parent.document;
+                void pDoc.title;
+                return pDoc;
+            } catch (err) { return null; }
+        };
+        const getThemeFromParentClass = () => {
+            try {
+                if (!window.parent || window.parent === window) return null;
+                const pDoc = window.parent.document;
+                const html = pDoc.documentElement;
+                const body = pDoc.body;
+                const htmlClass = html ? html.className : '';
+                const bodyClass = body ? body.className : '';
+                const htmlDataTheme = html ? html.getAttribute('data-theme') : '';
+                if (htmlDataTheme === 'dark' || bodyClass.includes('dark') || htmlClass.includes('dark')) return 'dark';
+                if (htmlDataTheme === 'light' || bodyClass.includes('light') || htmlClass.includes('light')) return 'light';
+                return null;
+            } catch (err) { return null; }
+        };
+        const setTheme = () => {
+            const parentDoc = getParentDocumentSafe();
+            const metaTheme = parentDoc ? getThemeFromMeta(parentDoc) : null;
+            const parentClassTheme = getThemeFromParentClass();
+            const prefersDark = window.matchMedia && window.matchMedia('(prefers-color-scheme: dark)').matches;
+            const chosen = metaTheme || parentClassTheme || (prefersDark ? 'dark' : 'light');
+            document.documentElement.classList.toggle('theme-dark', chosen === 'dark');
+        };
+        setTheme();
+        if (window.matchMedia) {
+            window.matchMedia('(prefers-color-scheme: dark)').addEventListener('change', setTheme);
+        }
+    })();
+    </script>
+</body>
+</html>
+"""
+
+# =================================================================
+# LLM 提示词 - 深度下潜思维链
+# =================================================================
+
+SYSTEM_PROMPT = """
+你是一位“深度下潜 (Deep Dive)”分析专家。你的目标是引导用户完成一个全面的思维过程，从表面理解深入到战略行动。
+
+## 思维结构 (严格遵守)
+
+你必须从以下四个维度剖析输入内容：
+
+### 1. 🔍 The Context (全景)
+提供一个高层级的全景视图。内容是关于什么的？核心情境、背景或正在解决的问题是什么？（2-3 段话）
+
+### 2. 🧠 The Logic (脉络)
+解构底层结构。论点是如何构建的？其中的推理逻辑、隐藏假设或起作用的思维模型是什么？（列表形式）
+
+### 3. 💎 The Insight (洞察)
+提取非显性的价值。有哪些“原来如此”的时刻？揭示了哪些深层含义、盲点或独特视角？（列表形式）
+
+### 4. 🚀 The Path (路径)
+定义战略方向。具体的、按优先级排列的下一步行动是什么？如何立即应用这些知识？（可执行步骤）
+
+## 规则
+- 使用用户指定的语言输出。
+- 保持专业、分析性且富有启发性的语调。
+- 聚焦于“理解的过程”，而不仅仅是结果。
+- 不要包含寒暄或元对话。
+"""
+
+USER_PROMPT = """
+对以下内容发起“深度下潜”：
+
+**用户上下文：**
+- 用户：{user_name}
+- 时间：{current_date_time_str}
+- 语言：{user_language}
+
+**待分析内容：**
+```
+{long_text_content}
+```
+
+请执行完整的思维链：全景 (Context) → 脉络 (Logic) → 洞察 (Insight) → 路径 (Path)。
+"""
+
+# =================================================================
+# 现代 CSS 设计 - 深度下潜主题
+# =================================================================
+
+CSS_TEMPLATE = """
+        .deep-dive {
+            font-family: 'Inter', -apple-system, system-ui, sans-serif;
+            color: var(--dd-text-secondary);
+        }
+
+        .dd-header {
+            background: var(--dd-header-gradient);
+            padding: 40px 32px;
+            color: white;
+            position: relative;
+        }
+
+        .dd-header-badge {
+            display: inline-block;
+            padding: 4px 12px;
+            background: rgba(255,255,255,0.1);
+            border: 1px solid rgba(255,255,255,0.2);
+            border-radius: 100px;
+            font-size: 0.75rem;
+            font-weight: 600;
+            letter-spacing: 0.05em;
+            text-transform: uppercase;
+            margin-bottom: 16px;
+        }
+
+        .dd-title {
+            font-size: 2rem;
+            font-weight: 800;
+            margin: 0 0 12px 0;
+            letter-spacing: -0.02em;
+        }
+
+        .dd-meta {
+            display: flex;
+            gap: 20px;
+            font-size: 0.85rem;
+            opacity: 0.7;
+        }
+
+        .dd-body {
+            padding: 32px;
+            display: flex;
+            flex-direction: column;
+            gap: 40px;
+            position: relative;
+            background: var(--dd-bg-primary);
+        }
+
+        /* 思维导火索 */
+        .dd-body::before {
+            content: '';
+            position: absolute;
+            left: 52px;
+            top: 40px;
+            bottom: 40px;
+            width: 2px;
+            background: var(--dd-border);
+            z-index: 0;
+        }
+
+        .dd-step {
+            position: relative;
+            z-index: 1;
+            display: flex;
+            gap: 24px;
+        }
+
+        .dd-step-icon {
+            flex-shrink: 0;
+            width: 40px;
+            height: 40px;
+            background: var(--dd-bg-primary);
+            border: 2px solid var(--dd-border);
+            border-radius: 12px;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            font-size: 1.25rem;
+            box-shadow: 0 4px 12px rgba(0,0,0,0.03);
+            transition: all 0.3s ease;
+        }
+
+        .dd-step:hover .dd-step-icon {
+            border-color: var(--dd-accent);
+            transform: scale(1.1);
+        }
+
+        .dd-step-content {
+            flex: 1;
+        }
+
+        .dd-step-label {
+            font-size: 0.75rem;
+            font-weight: 700;
+            color: var(--dd-accent);
+            text-transform: uppercase;
+            letter-spacing: 0.1em;
+            margin-bottom: 4px;
+        }
+
+        .dd-step-title {
+            font-size: 1.25rem;
+            font-weight: 700;
+            color: var(--dd-text-primary);
+            margin: 0 0 16px 0;
+        }
+
+        .dd-text {
+            line-height: 1.7;
+            font-size: 1rem;
+        }
+
+        .dd-text p { margin-bottom: 16px; }
+        .dd-text p:last-child { margin-bottom: 0; }
+
+        .dd-list {
+            list-style: none;
+            padding: 0;
+            margin: 0;
+            display: grid;
+            gap: 12px;
+        }
+
+        .dd-list-item {
+            background: var(--dd-bg-secondary);
+            padding: 16px 20px;
+            border-radius: 12px;
+            border-left: 4px solid var(--dd-border);
+            transition: all 0.2s ease;
+        }
+
+        .dd-list-item:hover {
+            background: var(--dd-bg-tertiary);
+            border-left-color: var(--dd-accent);
+            transform: translateX(4px);
+        }
+
+        .dd-list-item strong {
+            color: var(--dd-text-primary);
+            display: block;
+            margin-bottom: 4px;
+        }
+
+        .dd-path-item {
+            background: var(--dd-accent-soft);
+            border-left-color: var(--dd-accent);
+        }
+
+        .dd-footer {
+            padding: 24px 32px;
+            background: var(--dd-bg-secondary);
+            border-top: 1px solid var(--dd-border);
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            font-size: 0.8rem;
+            color: var(--dd-text-dim);
+        }
+
+        .dd-tag {
+            padding: 2px 8px;
+            background: var(--dd-bg-tertiary);
+            border-radius: 4px;
+            font-weight: 600;
+        }
+
+        .dd-text code,
+        .dd-list-item code {
+            background: var(--dd-code-bg);
+            color: var(--dd-text-primary);
+            padding: 2px 6px;
+            border-radius: 4px;
+            font-family: 'SF Mono', 'Consolas', 'Monaco', monospace;
+            font-size: 0.85em;
+        }
+
+        .dd-list-item em {
+            font-style: italic;
+            color: var(--dd-text-dim);
+        }
+"""
+
+CONTENT_TEMPLATE = """
+        <div class="deep-dive">
+            <div class="dd-header">
+                <div class="dd-header-badge">思维过程</div>
+                <h1 class="dd-title">精读分析报告</h1>
+                <div class="dd-meta">
+                    <span>👤 {user_name}</span>
+                    <span>📅 {current_date_time_str}</span>
+                    <span>📊 {word_count} 字</span>
+                </div>
+            </div>
+            <div class="dd-body">
+                <!-- 第一步：全景 -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🔍</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 01</div>
+                        <h2 class="dd-step-title">全景 (The Context)</h2>
+                        <div class="dd-text">{context_html}</div>
+                    </div>
+                </div>
+
+                <!-- 第二步：脉络 -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🧠</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 02</div>
+                        <h2 class="dd-step-title">脉络 (The Logic)</h2>
+                        <div class="dd-text">{logic_html}</div>
+                    </div>
+                </div>
+
+                <!-- 第三步：洞察 -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">💎</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 03</div>
+                        <h2 class="dd-step-title">洞察 (The Insight)</h2>
+                        <div class="dd-text">{insight_html}</div>
+                    </div>
+                </div>
+
+                <!-- 第四步：路径 -->
+                <div class="dd-step">
+                    <div class="dd-step-icon">🚀</div>
+                    <div class="dd-step-content">
+                        <div class="dd-step-label">Phase 04</div>
+                        <h2 class="dd-step-title">路径 (The Path)</h2>
+                        <div class="dd-text">{path_html}</div>
+                    </div>
+                </div>
+            </div>
+            <div class="dd-footer">
+                <span>Deep Dive Engine v1.0</span>
+                <span><span class="dd-tag">AI 驱动分析</span></span>
+            </div>
+        </div>
+"""
+
+
+class Action:
+    class Valves(BaseModel):
+        SHOW_STATUS: bool = Field(
+            default=True,
+            description="是否显示操作状态更新。",
+        )
+        MODEL_ID: str = Field(
+            default="",
+            description="用于分析的 LLM 模型 ID。留空则使用当前模型。",
+        )
+        MIN_TEXT_LENGTH: int = Field(
+            default=200,
+            description="深度下潜所需的最小文本长度（字符）。",
+        )
+        CLEAR_PREVIOUS_HTML: bool = Field(
+            default=True,
+            description="是否清除之前的插件结果。",
+        )
+        MESSAGE_COUNT: int = Field(
+            default=1,
+            description="要分析的最近消息数量。",
+        )
+
+    def __init__(self):
+        self.valves = self.Valves()
+
+    def _get_user_context(self, __user__: Optional[Dict[str, Any]]) -> Dict[str, str]:
+        """安全提取用户上下文信息。"""
+        if isinstance(__user__, (list, tuple)):
+            user_data = __user__[0] if __user__ else {}
+        elif isinstance(__user__, dict):
+            user_data = __user__
+        else:
+            user_data = {}
+
+        return {
+            "user_id": user_data.get("id", "unknown_user"),
+            "user_name": user_data.get("name", "用户"),
+            "user_language": user_data.get("language", "zh-CN"),
+        }
+
+    def _process_llm_output(self, llm_output: str) -> Dict[str, str]:
+        """解析 LLM 输出并转换为样式化 HTML。"""
+        # 使用灵活的正则提取各部分
+        context_match = re.search(
+            r"###\s*1\.\s*🔍?\s*(?:全景|The Context)\s*(?:\((.*?)\))?\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        logic_match = re.search(
+            r"###\s*2\.\s*🧠?\s*(?:脉络|The Logic)\s*(?:\((.*?)\))?\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        insight_match = re.search(
+            r"###\s*3\.\s*💎?\s*(?:洞察|The Insight)\s*(?:\((.*?)\))?\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+        path_match = re.search(
+            r"###\s*4\.\s*🚀?\s*(?:路径|The Path)\s*(?:\((.*?)\))?\s*\n(.*?)(?=\n###|$)",
+            llm_output,
+            re.DOTALL | re.IGNORECASE,
+        )
+
+        # 兜底正则
+        if not context_match:
+            context_match = re.search(
+                r"###\s*🔍?\s*(?:全景|The Context).*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not logic_match:
+            logic_match = re.search(
+                r"###\s*🧠?\s*(?:脉络|The Logic).*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not insight_match:
+            insight_match = re.search(
+                r"###\s*💎?\s*(?:洞察|The Insight).*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+        if not path_match:
+            path_match = re.search(
+                r"###\s*🚀?\s*(?:路径|The Path).*?\n(.*?)(?=\n###|$)",
+                llm_output,
+                re.DOTALL | re.IGNORECASE,
+            )
+
+        context_md = (
+            context_match.group(context_match.lastindex).strip()
+            if context_match
+            else ""
+        )
+        logic_md = (
+            logic_match.group(logic_match.lastindex).strip() if logic_match else ""
+        )
+        insight_md = (
+            insight_match.group(insight_match.lastindex).strip()
+            if insight_match
+            else ""
+        )
+        path_md = path_match.group(path_match.lastindex).strip() if path_match else ""
+
+        if not any([context_md, logic_md, insight_md, path_md]):
+            context_md = llm_output.strip()
+            logger.warning("LLM 输出未遵循格式，将作为全景处理。")
+
+        md_extensions = ["nl2br"]
+
+        context_html = (
+            markdown.markdown(context_md, extensions=md_extensions)
+            if context_md
+            else '<p class="dd-no-content">未能提取全景信息。</p>'
+        )
+        logic_html = (
+            self._process_list_items(logic_md, "logic")
+            if logic_md
+            else '<p class="dd-no-content">未能解构脉络。</p>'
+        )
+        insight_html = (
+            self._process_list_items(insight_md, "insight")
+            if insight_md
+            else '<p class="dd-no-content">未能发现洞察。</p>'
+        )
+        path_html = (
+            self._process_list_items(path_md, "path")
+            if path_md
+            else '<p class="dd-no-content">未能定义路径。</p>'
+        )
+
+        return {
+            "context_html": context_html,
+            "logic_html": logic_html,
+            "insight_html": insight_html,
+            "path_html": path_html,
+        }
+
+    def _process_list_items(self, md_content: str, section_type: str) -> str:
+        """将 markdown 列表转换为样式化卡片，支持完整的 markdown 格式。"""
+        lines = md_content.strip().split("\n")
+        items = []
+        current_paragraph = []
+
+        for line in lines:
+            line = line.strip()
+
+            # 检查列表项（无序或有序）
+            bullet_match = re.match(r"^[-*]\s+(.+)$", line)
+            numbered_match = re.match(r"^\d+\.\s+(.+)$", line)
+
+            if bullet_match or numbered_match:
+                # 清空累积的段落
+                if current_paragraph:
+                    para_text = " ".join(current_paragraph)
+                    para_html = self._convert_inline_markdown(para_text)
+                    items.append(f"<p>{para_html}</p>")
+                    current_paragraph = []
+
+                # 提取列表项内容
+                text = (
+                    bullet_match.group(1) if bullet_match else numbered_match.group(1)
+                )
+
+                # 处理粗体标题模式：**标题:** 描述 或 **标题**: 描述
+                title_match = re.match(r"\*\*(.+?)\*\*[:\s：]*(.*)$", text)
+                if title_match:
+                    title = self._convert_inline_markdown(title_match.group(1))
+                    desc = self._convert_inline_markdown(title_match.group(2).strip())
+                    path_class = "dd-path-item" if section_type == "path" else ""
+                    item_html = f'<div class="dd-list-item {path_class}"><strong>{title}</strong>{desc}</div>'
+                else:
+                    text_html = self._convert_inline_markdown(text)
+                    path_class = "dd-path-item" if section_type == "path" else ""
+                    item_html = (
+                        f'<div class="dd-list-item {path_class}">{text_html}</div>'
+                    )
+                items.append(item_html)
+            elif line and not line.startswith("#"):
+                # 累积段落文本
+                current_paragraph.append(line)
+            elif not line and current_paragraph:
+                # 空行结束段落
+                para_text = " ".join(current_paragraph)
+                para_html = self._convert_inline_markdown(para_text)
+                items.append(f"<p>{para_html}</p>")
+                current_paragraph = []
+
+        # 清空剩余段落
+        if current_paragraph:
+            para_text = " ".join(current_paragraph)
+            para_html = self._convert_inline_markdown(para_text)
+            items.append(f"<p>{para_html}</p>")
+
+        if items:
+            return f'<div class="dd-list">{" ".join(items)}</div>'
+        return f'<p class="dd-no-content">未找到条目。</p>'
+
+    def _convert_inline_markdown(self, text: str) -> str:
+        """将行内 markdown（粗体、斜体、代码）转换为 HTML。"""
+        # 转换行内代码：`code` -> <code>code</code>
+        text = re.sub(r"`([^`]+)`", r"<code>\1</code>", text)
+        # 转换粗体：**text** -> <strong>text</strong>
+        text = re.sub(r"\*\*(.+?)\*\*", r"<strong>\1</strong>", text)
+        # 转换斜体：*text* -> <em>text</em>（但不在 ** 内部）
+        text = re.sub(r"(?<!\*)\*([^*]+)\*(?!\*)", r"<em>\1</em>", text)
+        return text
+
+    async def _emit_status(
+        self,
+        emitter: Optional[Callable[[Any], Awaitable[None]]],
+        description: str,
+        done: bool = False,
+    ):
+        """发送状态更新事件。"""
+        if self.valves.SHOW_STATUS and emitter:
+            await emitter(
+                {"type": "status", "data": {"description": description, "done": done}}
+            )
+
+    async def _emit_notification(
+        self,
+        emitter: Optional[Callable[[Any], Awaitable[None]]],
+        content: str,
+        ntype: str = "info",
+    ):
+        """发送通知事件。"""
+        if emitter:
+            await emitter(
+                {"type": "notification", "data": {"type": ntype, "content": content}}
+            )
+
+    def _remove_existing_html(self, content: str) -> str:
+        """移除已有的插件生成的 HTML。"""
+        pattern = r"```html\s*<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?```"
+        return re.sub(pattern, "", content).strip()
+
+    def _extract_text_content(self, content) -> str:
+        """从消息内容中提取文本。"""
+        if isinstance(content, str):
+            return content
+        elif isinstance(content, list):
+            text_parts = []
+            for item in content:
+                if isinstance(item, dict) and item.get("type") == "text":
+                    text_parts.append(item.get("text", ""))
+                elif isinstance(item, str):
+                    text_parts.append(item)
+            return "\n".join(text_parts)
+        return str(content) if content else ""
+
+    def _merge_html(
+        self,
+        existing_html: str,
+        new_content: str,
+        new_styles: str = "",
+        user_language: str = "zh-CN",
+    ) -> str:
+        """合并新内容到 HTML 容器。"""
+        if "<!-- OPENWEBUI_PLUGIN_OUTPUT -->" in existing_html:
+            base_html = re.sub(r"^```html\s*", "", existing_html)
+            base_html = re.sub(r"\s*```$", "", base_html)
+        else:
+            base_html = HTML_WRAPPER_TEMPLATE.replace("{user_language}", user_language)
+
+        wrapped = f'<div class="plugin-item">\n{new_content}\n</div>'
+
+        if new_styles:
+            base_html = base_html.replace(
+                "/* STYLES_INSERTION_POINT */",
+                f"{new_styles}\n/* STYLES_INSERTION_POINT */",
+            )
+
+        base_html = base_html.replace(
+            "<!-- CONTENT_INSERTION_POINT -->",
+            f"{wrapped}\n<!-- CONTENT_INSERTION_POINT -->",
+        )
+
+        return base_html.strip()
+
+    def _build_content_html(self, context: dict) -> str:
+        """构建内容 HTML。"""
+        html = CONTENT_TEMPLATE
+        for key, value in context.items():
+            html = html.replace(f"{{{key}}}", str(value))
+        return html
+
+    async def action(
+        self,
+        body: dict,
+        __user__: Optional[Dict[str, Any]] = None,
+        __event_emitter__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __request__: Optional[Request] = None,
+    ) -> Optional[dict]:
+        logger.info("Action: 精读 v1.0.0 启动")
+
+        user_ctx = self._get_user_context(__user__)
+        user_id = user_ctx["user_id"]
+        user_name = user_ctx["user_name"]
+        user_language = user_ctx["user_language"]
+
+        now = datetime.now()
+        current_date_time_str = now.strftime("%Y年%m月%d日 %H:%M")
+
+        original_content = ""
+        try:
+            messages = body.get("messages", [])
+            if not messages:
+                raise ValueError("未找到消息内容。")
+
+            message_count = min(self.valves.MESSAGE_COUNT, len(messages))
+            recent_messages = messages[-message_count:]
+
+            aggregated_parts = []
+            for msg in recent_messages:
+                text = self._extract_text_content(msg.get("content"))
+                if text:
+                    aggregated_parts.append(text)
+
+            if not aggregated_parts:
+                raise ValueError("未找到文本内容。")
+
+            original_content = "\n\n---\n\n".join(aggregated_parts)
+            word_count = len(original_content)
+
+            if len(original_content) < self.valves.MIN_TEXT_LENGTH:
+                msg = f"内容过短（{len(original_content)} 字符）。精读至少需要 {self.valves.MIN_TEXT_LENGTH} 字符才能进行有意义的分析。"
+                await self._emit_notification(__event_emitter__, msg, "warning")
+                return {"messages": [{"role": "assistant", "content": f"⚠️ {msg}"}]}
+
+            await self._emit_notification(
+                __event_emitter__, "📖 正在发起精读分析...", "info"
+            )
+            await self._emit_status(
+                __event_emitter__, "📖 精读：正在分析全景与脉络...", False
+            )
+
+            prompt = USER_PROMPT.format(
+                user_name=user_name,
+                current_date_time_str=current_date_time_str,
+                user_language=user_language,
+                long_text_content=original_content,
+            )
+
+            model = self.valves.MODEL_ID or body.get("model")
+            payload = {
+                "model": model,
+                "messages": [
+                    {"role": "system", "content": SYSTEM_PROMPT},
+                    {"role": "user", "content": prompt},
+                ],
+                "stream": False,
+            }
+
+            user_obj = Users.get_user_by_id(user_id)
+            if not user_obj:
+                raise ValueError(f"未找到用户：{user_id}")
+
+            response = await generate_chat_completion(__request__, payload, user_obj)
+            llm_output = response["choices"][0]["message"]["content"]
+
+            processed = self._process_llm_output(llm_output)
+
+            context = {
+                "user_name": user_name,
+                "current_date_time_str": current_date_time_str,
+                "word_count": word_count,
+                **processed,
+            }
+
+            content_html = self._build_content_html(context)
+
+            # 处理已有 HTML
+            existing = ""
+            match = re.search(
+                r"```html\s*(<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?)```",
+                original_content,
+            )
+            if match:
+                existing = match.group(1)
+
+            if self.valves.CLEAR_PREVIOUS_HTML or not existing:
+                original_content = self._remove_existing_html(original_content)
+                final_html = self._merge_html(
+                    "", content_html, CSS_TEMPLATE, user_language
+                )
+            else:
+                original_content = self._remove_existing_html(original_content)
+                final_html = self._merge_html(
+                    existing, content_html, CSS_TEMPLATE, user_language
+                )
+
+            body["messages"][-1][
+                "content"
+            ] = f"{original_content}\n\n```html\n{final_html}\n```"
+
+            await self._emit_status(__event_emitter__, "📖 精读完成！", True)
+            await self._emit_notification(
+                __event_emitter__,
+                f"📖 精读完成，{user_name}！思维链已生成。",
+                "success",
+            )
+
+        except Exception as e:
+            logger.error(f"Deep Dive 错误：{e}", exc_info=True)
+            body["messages"][-1][
+                "content"
+            ] = f"{original_content}\n\n❌ **错误：** {str(e)}"
+            await self._emit_status(__event_emitter__, "精读失败。", True)
+            await self._emit_notification(__event_emitter__, f"错误：{str(e)}", "error")
+
+        return body

plugins/actions/export_to_docx/README.md

@@ -1,74 +1,88 @@
-# Export to Word
+# 📝 Export to Word (Enhanced)
 
-Export current conversation from Markdown to Word (.docx) with **syntax highlighting**, **blockquote support**, and smarter filenames.
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.4.3 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
 
-## Features
+Export conversation to Word (.docx) with **syntax highlighting**, **native math equations**, **Mermaid diagrams**, **citations**, and **enhanced table formatting**.
 
-- **One-Click Export**: Adds an "Export to Word" action button to the chat.
-- **Markdown Conversion**: Converts Markdown syntax to Word formatting (headings, bold, italic, code, tables, lists).
-- **Syntax Highlighting**: Code blocks are highlighted with Pygments (supports 500+ languages).
-- **Blockquote Support**: Markdown blockquotes are rendered with left border and gray styling.
-- **Multi-language Support**: Properly handles both Chinese and English text without garbled characters.
-- **Smarter Filenames**: Configurable title source (Chat Title, AI Generated, or Markdown Title).
+## 🔥 What's New in v0.4.3
 
-## Configuration
+- ✨ **S3 Object Storage Support**: Direct access to images stored in S3/MinIO via boto3, bypassing API layer for faster exports.
+- 🔧 **Multi-level File Fallback**: 6-level fallback mechanism for file retrieval (DB → S3 → Local → URL → API → Attributes).
+- 🛡️ **Improved Error Handling**: Better logging and error messages for file retrieval failures.
 
-You can configure the following settings via the **Valves** button in the plugin settings:
+## ✨ Key Features
 
-- **TITLE_SOURCE**: Choose how the document title/filename is generated.
-    - `chat_title`: Use the conversation title (default).
-    - `ai_generated`: Use AI to generate a short title based on the content.
-    - `markdown_title`: Extract the first h1/h2 heading from the Markdown content.
+- 🚀 **One-Click Export**: Adds an "Export to Word" action button to the chat.
+- 📄 **Markdown Conversion**: Full Markdown syntax support (headings, bold, italic, code, tables, lists).
+- 🎨 **Syntax Highlighting**: Code blocks highlighted with Pygments (500+ languages).
+- 🔢 **Native Math Equations**: LaTeX math (`$$...$$`, `\[...\]`, `$...$`) converted to editable Word equations.
+- 📊 **Mermaid Diagrams**: Flowcharts and sequence diagrams rendered as images.
+- 📚 **Citations & References**: Auto-generates References section with clickable citation links.
+- 🧹 **Reasoning Stripping**: Automatically removes AI thinking blocks (`<think>`, `<analysis>`).
+- 📋 **Enhanced Tables**: Smart column widths, alignment, header row repeat across pages.
+- 💬 **Blockquote Support**: Markdown blockquotes with left border and gray styling.
+- 🌐 **Multi-language Support**: Proper handling of Chinese and English text.
 
-## Supported Markdown Syntax
+## 🚀 How to Use
+
+1. **Install**: Search for "Export to Word" in the Open WebUI Community and install.
+2. **Trigger**: In any chat, click the "Export to Word" action button.
+3. **Download**: The .docx file will be automatically downloaded.
+
+## ⚙️ Configuration (Valves)
+
+| Parameter | Default | Description |
+| :--- | :--- | :--- |
+| **Title Source (TITLE_SOURCE)** | `chat_title` | `chat_title`, `ai_generated`, or `markdown_title` |
+| **Max Image Size (MAX_EMBED_IMAGE_MB)** | `20` | Maximum image size to embed (MB) |
+| **UI Language (UI_LANGUAGE)** | `en` | `en` (English) or `zh` (Chinese) |
+| **Latin Font (FONT_LATIN)** | `Times New Roman` | Font for Latin characters |
+| **Asian Font (FONT_ASIAN)** | `SimSun` | Font for Asian characters |
+| **Code Font (FONT_CODE)** | `Consolas` | Font for code blocks |
+| **Table Header Color** | `F2F2F2` | Header background color (hex) |
+| **Table Zebra Color** | `FBFBFB` | Alternating row color (hex) |
+| **Mermaid PNG Scale** | `3.0` | Resolution multiplier for Mermaid images |
+| **Math Enable** | `True` | Enable LaTeX math conversion |
+
+## 🛠️ Supported Markdown Syntax
 
 | Syntax | Word Result |
-| :---------------------------------- | :-------------------------------- |
+| :--- | :--- |
 | `# Heading 1` to `###### Heading 6` | Heading levels 1-6 |
 | `**bold**` or `__bold__` | Bold text |
 | `*italic*` or `_italic_` | Italic text |
-| `***bold italic***`                 | Bold + Italic                     |
 | `` `inline code` `` | Monospace with gray background |
 | ` ``` code block ``` ` | **Syntax highlighted** code block |
 | `> blockquote` | Left-bordered gray italic text |
-| `[link](url)`                       | Blue underlined link text         |
+| `[link](url)` | Blue underlined link |
 | `~~strikethrough~~` | Strikethrough text |
 | `- item` or `* item` | Bullet list |
 | `1. item` | Numbered list |
-| Markdown tables                     | Table with grid                   |
-| `---` or `***`                      | Horizontal rule                   |
+| Markdown tables | **Enhanced table** with smart widths |
+| `$$LaTeX$$` or `\[LaTeX\]` | **Native Word equation** (display) |
+| `$LaTeX$` or `\(LaTeX\)` | **Native Word equation** (inline) |
+| ` ```mermaid ... ``` ` | **Mermaid diagram** as image |
+| `[1]` citation markers | **Clickable links** to References |
 
-## Usage
-
-1. Install the plugin.
-2. In any chat, click the "Export to Word" button.
-3. The .docx file will be automatically downloaded to your device.
-
-
-### Notes
-
-- Title detection only considers h1/h2 headings.
-- If the request carries `chat_id` (body or metadata), the plugin will fetch the chat title from the database when the body lacks one.
-- Default fonts: Times New Roman (en), SimSun/SimHei (zh), Consolas (code).
-
-### Requirements
+## 📦 Requirements
 
 - `python-docx==1.1.2` - Word document generation
-- `Pygments>=2.15.0` - Syntax highlighting (optional but recommended)
+- `Pygments>=2.15.0` - Syntax highlighting
+- `latex2mathml` - LaTeX to MathML conversion
+- `mathml2omml` - MathML to Office Math (OMML) conversion
 
-Both are declared in the plugin docstring; ensure they are installed in your environment.
+## 📝 Changelog
 
-## Font Configuration
+### v0.4.3
+- **S3 Object Storage**: Direct S3/MinIO access via boto3 for faster image retrieval.
+- **6-Level Fallback**: Robust file retrieval: DB → S3 → Local → URL → API → Attributes.
+- **Better Logging**: Improved error messages for debugging file access issues.
 
-- **English Text**: Times New Roman
-- **Chinese Text**: SimSun (宋体) for body, SimHei (黑体) for headings
-- **Code**: Consolas
+### v0.4.1
+- **Chinese Parameter Names**: Localized configuration names for Chinese version.
 
-## Author
-
-Fu-Jie  
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## License
-
-MIT License
+### v0.4.0
+- **Multi-language Support**: UI language switching (English/Chinese).
+- **Font & Style Configuration**: Customizable fonts and table colors.
+- **Mermaid Enhancements**: Hybrid SVG+PNG rendering, background color config.
+- **Performance**: Real-time progress updates for large exports.

plugins/actions/export_to_docx/README_CN.md

@@ -1,33 +1,56 @@
-# 导出为 Word
+# 📝 导出为 Word (增强版)
 
-将当前对话内容从 Markdown 转换并导出为 Word (.docx) 文件，支持**代码语法高亮**、**引用块样式**和更智能的文件命名。
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.4.3 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
 
-## 功能特点
+将对话导出为 Word (.docx)，支持**代码语法高亮**、**原生数学公式**、**Mermaid 图表**、**引用参考**和**增强表格格式**。
 
-- **一键导出**：在聊天界面添加“导出为 Word”动作按钮。
-- **Markdown 转换**：将 Markdown 语法转换为 Word 格式（标题、粗体、斜体、代码、表格、列表）。
-- **代码语法高亮**：使用 Pygments 库为代码块添加语法高亮（支持 500+ 种语言）。
-- **引用块支持**：Markdown 引用块会渲染为带左侧边框的灰色斜体样式。
-- **多语言支持**：正确处理中文和英文文本，无乱码问题。
-- **更智能的文件名**：可配置标题来源（对话标题、AI 生成或 Markdown 标题）。
+## 🔥 v0.4.3 更新内容
 
-## 配置 (Configuration)
+- ✨ **S3 对象存储支持**: 通过 boto3 直连 S3/MinIO，绕过 API 层，导出速度更快。
+- 🔧 **多级文件回退**: 6 级文件获取机制（数据库 → S3 → 本地 → URL → API → 属性）。
+- 🛡️ **错误处理优化**: 更完善的日志记录和错误提示，便于调试文件访问问题。
 
-您可以通过插件设置中的 **Valves** 按钮配置以下选项：
+## ✨ 核心特性
 
-- **TITLE_SOURCE**：选择文档标题/文件名的生成方式。
-    - `chat_title`：使用对话标题（默认）。
-    - `ai_generated`：使用 AI 根据内容生成简短标题。
-    - `markdown_title`：从 Markdown 内容中提取第一个一级或二级标题。
+- 🚀 **一键导出**: 在聊天界面添加"导出为 Word"动作按钮。
+- 📄 **Markdown 转换**: 完整支持 Markdown 语法（标题、粗体、斜体、代码、表格、列表）。
+- 🎨 **代码语法高亮**: 使用 Pygments 库高亮代码块（支持 500+ 种语言）。
+- 🔢 **原生数学公式**: LaTeX 公式（`$$...$$`、`\[...\]`、`$...$`）转换为可编辑的 Word 公式。
+- 📊 **Mermaid 图表**: 流程图和时序图渲染为文档中的图片。
+- 📚 **引用与参考**: 自动生成参考资料章节，支持可点击的引用链接。
+- 🧹 **移除思考过程**: 自动移除 AI 思考块（`<think>`、`<analysis>`）。
+- 📋 **增强表格**: 智能列宽、对齐、表头跨页重复。
+- 💬 **引用块支持**: Markdown 引用块渲染为带左侧边框的灰色斜体样式。
+- 🌐 **多语言支持**: 正确处理中文和英文文本。
 
-## 支持的 Markdown 语法
+## 🚀 使用方法
+
+1. **安装**: 在 Open WebUI 社区搜索 "导出为 Word" 并安装。
+2. **触发**: 在任意对话中，点击"导出为 Word"动作按钮。
+3. **下载**: .docx 文件将自动下载到你的设备。
+
+## ⚙️ 配置参数 (Valves)
+
+| 参数 | 默认值 | 说明 |
+| :--- | :--- | :--- |
+| **文档标题来源** | `chat_title` | `chat_title`（对话标题）、`ai_generated`（AI 生成）、`markdown_title`（Markdown 标题）|
+| **最大嵌入图片大小MB** | `20` | 嵌入图片的最大大小 (MB) |
+| **界面语言** | `zh` | `en`（英语）或 `zh`（中文）|
+| **英文字体** | `Calibri` | 英文字体名称 |
+| **中文字体** | `SimSun` | 中文字体名称 |
+| **代码字体** | `Consolas` | 代码块字体名称 |
+| **表头背景色** | `F2F2F2` | 表头背景色（十六进制）|
+| **表格隔行背景色** | `FBFBFB` | 表格隔行背景色（十六进制）|
+| **Mermaid_PNG缩放比例** | `3.0` | Mermaid 图片分辨率倍数 |
+| **启用数学公式** | `True` | 启用 LaTeX 公式转换 |
+
+## 🛠️ 支持的 Markdown 语法
 
 | 语法 | Word 效果 |
-| :-------------------------- | :----------------------- |
+| :--- | :--- |
 | `# 标题1` 到 `###### 标题6` | 标题级别 1-6 |
 | `**粗体**` 或 `__粗体__` | 粗体文本 |
 | `*斜体*` 或 `_斜体_` | 斜体文本 |
-| `***粗斜体***`              | 粗体 + 斜体              |
 | `` `行内代码` `` | 等宽字体 + 灰色背景 |
 | ` ``` 代码块 ``` ` | **语法高亮**的代码块 |
 | `> 引用文本` | 带左侧边框的灰色斜体文本 |
@@ -35,39 +58,31 @@
 | `~~删除线~~` | 删除线文本 |
 | `- 项目` 或 `* 项目` | 无序列表 |
 | `1. 项目` | 有序列表 |
-| Markdown 表格               | 带边框表格               |
-| `---` 或 `***`              | 水平分割线               |
+| Markdown 表格 | **增强表格**（智能列宽）|
+| `$$LaTeX$$` 或 `\[LaTeX\]` | **原生 Word 公式**（块级）|
+| `$LaTeX$` 或 `\(LaTeX\)` | **原生 Word 公式**（行内）|
+| ` ```mermaid ... ``` ` | **Mermaid 图表**（图片形式）|
+| `[1]` 引用标记 | **可点击链接**到参考资料 |
 
-## 使用方法
-
-1. 安装插件。
-2. 在任意对话中，点击"导出为 Word"按钮。
-3. .docx 文件将自动下载到你的设备。
-
-### 说明
-
-- 标题检测仅考虑一级/二级标题（h1/h2）。
-- 若请求体或 metadata 提供 `chat_id`，当正文缺少标题时会从数据库查询对话标题。
-- 默认字体：英文 Times New Roman，中文宋体/黑体，代码 Consolas。
-
-### 依赖
+## 📦 依赖
 
 - `python-docx==1.1.2` - Word 文档生成
-- `Pygments>=2.15.0` - 语法高亮（可选但建议安装）
+- `Pygments>=2.15.0` - 语法高亮
+- `latex2mathml` - LaTeX 转 MathML
+- `mathml2omml` - MathML 转 Office Math (OMML)
 
-两者已在插件文档字符串中声明，请确保环境已安装。
+## 📝 更新日志
 
-## 字体配置
+### v0.4.3
+- **S3 对象存储**: 通过 boto3 直连 S3/MinIO，图片获取速度更快。
+- **6 级回退机制**: 稳健的文件获取：数据库 → S3 → 本地 → URL → API → 属性。
+- **日志优化**: 改进错误提示，便于调试文件访问问题。
 
-- **英文文本**：Times New Roman
-- **中文文本**：宋体（正文）、黑体（标题）
-- **代码**：Consolas
+### v0.4.1
+- **中文参数名**: 配置项名称和描述全部汉化。
 
-## 作者
-
-Fu-Jie  
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## 许可证
-
-MIT License
+### v0.4.0
+- **多语言支持**: 界面语言切换（中文/英文）。
+- **字体与样式配置**: 支持自定义中英文字体、代码字体以及表格颜色。
+- **Mermaid 增强**: 混合 SVG+PNG 渲染，支持背景色配置。
+- **性能优化**: 导出大型文档时提供实时进度反馈。

plugins/actions/export_to_docx/导出为Word.py

@@ -1,882 +0,0 @@
-"""
-title: 导出为 Word
-author: Fu-Jie
-author_url: https://github.com/Fu-Jie
-funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.1.0
-icon_url: data:image/svg+xml;base64,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
-requirements: python-docx==1.1.2, Pygments>=2.15.0
-description: 将当前对话内容从 Markdown 转换并导出为 Word (.docx) 文件，支持代码语法高亮和引用块。
-"""
-
-import os
-import re
-import base64
-import datetime
-import io
-import asyncio
-import logging
-from typing import Optional, Callable, Awaitable, Any, List, Tuple
-from docx import Document
-from docx.shared import Pt, Inches, RGBColor, Cm
-from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_LINE_SPACING
-from docx.enum.table import WD_TABLE_ALIGNMENT
-from docx.enum.style import WD_STYLE_TYPE
-from docx.oxml.ns import qn
-from docx.oxml import OxmlElement
-from open_webui.models.chats import Chats
-from open_webui.models.users import Users
-from open_webui.utils.chat import generate_chat_completion
-from pydantic import BaseModel, Field
-
-# Pygments for syntax highlighting
-try:
-    from pygments import lex
-    from pygments.lexers import get_lexer_by_name, TextLexer
-    from pygments.token import Token
-
-    PYGMENTS_AVAILABLE = True
-except ImportError:
-    PYGMENTS_AVAILABLE = False
-
-
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
-)
-logger = logging.getLogger(__name__)
-
-
-class Action:
-    class Valves(BaseModel):
-        TITLE_SOURCE: str = Field(
-            default="chat_title",
-            description="标题来源: 'chat_title' (对话标题), 'ai_generated' (AI 生成), 'markdown_title' (Markdown 标题)",
-        )
-
-    def __init__(self):
-        self.valves = self.Valves()
-
-    async def _send_notification(self, emitter: Callable, type: str, content: str):
-        await emitter(
-            {"type": "notification", "data": {"type": type, "content": content}}
-        )
-
-    async def action(
-        self,
-        body: dict,
-        __user__=None,
-        __event_emitter__=None,
-        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
-        __metadata__: Optional[dict] = None,
-        __request__: Optional[Any] = None,
-    ):
-        logger.info(f"action:{__name__}")
-
-        # 解析用户信息
-        if isinstance(__user__, (list, tuple)):
-            user_language = (
-                __user__[0].get("language", "zh-CN") if __user__ else "zh-CN"
-            )
-            user_name = __user__[0].get("name", "用户") if __user__[0] else "用户"
-            user_id = (
-                __user__[0]["id"]
-                if __user__ and "id" in __user__[0]
-                else "unknown_user"
-            )
-        elif isinstance(__user__, dict):
-            user_language = __user__.get("language", "zh-CN")
-            user_name = __user__.get("name", "用户")
-            user_id = __user__.get("id", "unknown_user")
-
-        if __event_emitter__:
-            last_assistant_message = body["messages"][-1]
-
-            await __event_emitter__(
-                {
-                    "type": "status",
-                    "data": {"description": "正在转换为 Word 文档...", "done": False},
-                }
-            )
-
-            try:
-                message_content = last_assistant_message["content"]
-
-                if not message_content or not message_content.strip():
-                    await self._send_notification(
-                        __event_emitter__, "error", "没有找到可导出的内容！"
-                    )
-                    return
-
-                # 生成文件名
-                title = ""
-                chat_id = self.extract_chat_id(body, __metadata__)
-
-                # 直接通过 chat_id 获取标题，因为 body 中通常不包含标题
-                chat_title = ""
-                if chat_id:
-                    chat_title = await self.fetch_chat_title(chat_id, user_id)
-
-                # 根据配置决定文件名使用的标题
-                if (
-                    self.valves.TITLE_SOURCE == "chat_title"
-                    or not self.valves.TITLE_SOURCE
-                ):
-                    title = chat_title
-                elif self.valves.TITLE_SOURCE == "markdown_title":
-                    title = self.extract_title(message_content)
-                elif self.valves.TITLE_SOURCE == "ai_generated":
-                    title = await self.generate_title_using_ai(
-                        body, message_content, user_id, __request__
-                    )
-
-                current_datetime = datetime.datetime.now()
-                formatted_date = current_datetime.strftime("%Y%m%d")
-
-                if title:
-                    filename = f"{self.clean_filename(title)}.docx"
-                else:
-                    filename = f"{user_name}_{formatted_date}.docx"
-
-                # 创建 Word 文档；若正文无一级标题，使用对话标题作为一级标题
-                # 如果选择了 chat_title 且获取到了，则作为 top_heading
-                # 如果选择了其他方式，title 就是文件名，也可以作为 top_heading
-
-                # 保持原有逻辑：top_heading 主要是为了在文档开头补充标题
-                # 这里我们尽量使用 chat_title 作为 top_heading，如果它存在的话，因为它通常是对话的主题
-                # 即使文件名是 AI 生成的，文档内的标题用 chat_title 也是合理的
-                # 但如果用户选择了 markdown_title，可能不希望插入 chat_title
-
-                top_heading = ""
-                if chat_title:
-                    top_heading = chat_title
-                elif title:
-                    top_heading = title
-
-                has_h1 = bool(re.search(r"^#\s+.+$", message_content, re.MULTILINE))
-                doc = self.markdown_to_docx(
-                    message_content, top_heading=top_heading, has_h1=has_h1
-                )
-
-                # 保存到内存
-                doc_buffer = io.BytesIO()
-                doc.save(doc_buffer)
-                doc_buffer.seek(0)
-                file_content = doc_buffer.read()
-                base64_blob = base64.b64encode(file_content).decode("utf-8")
-
-                # 触发文件下载
-                if __event_call__:
-                    await __event_call__(
-                        {
-                            "type": "execute",
-                            "data": {
-                                "code": f"""
-                                try {{
-                                    const base64Data = "{base64_blob}";
-                                    const binaryData = atob(base64Data);
-                                    const arrayBuffer = new Uint8Array(binaryData.length);
-                                    for (let i = 0; i < binaryData.length; i++) {{
-                                        arrayBuffer[i] = binaryData.charCodeAt(i);
-                                    }}
-                                    const blob = new Blob([arrayBuffer], {{ type: "application/vnd.openxmlformats-officedocument.wordprocessingml.document" }});
-                                    const filename = "{filename}";
-
-                                    const url = URL.createObjectURL(blob);
-                                    const a = document.createElement("a");
-                                    a.style.display = "none";
-                                    a.href = url;
-                                    a.download = filename;
-                                    document.body.appendChild(a);
-                                    a.click();
-                                    URL.revokeObjectURL(url);
-                                    document.body.removeChild(a);
-                                }} catch (error) {{
-                                    console.error('触发下载时出错:', error);
-                                }}
-                                """
-                            },
-                        }
-                    )
-
-                await __event_emitter__(
-                    {
-                        "type": "status",
-                        "data": {"description": "Word 文档已导出", "done": True},
-                    }
-                )
-
-                await self._send_notification(
-                    __event_emitter__, "success", f"已成功导出为 {filename}"
-                )
-
-                return {"message": "下载事件已触发"}
-
-            except Exception as e:
-                print(f"Error exporting to Word: {str(e)}")
-                await __event_emitter__(
-                    {
-                        "type": "status",
-                        "data": {
-                            "description": f"导出失败: {str(e)}",
-                            "done": True,
-                        },
-                    }
-                )
-                await self._send_notification(
-                    __event_emitter__, "error", f"导出 Word 文档时出错: {str(e)}"
-                )
-
-    async def generate_title_using_ai(
-        self, body: dict, content: str, user_id: str, request: Any
-    ) -> str:
-        if not request:
-            return ""
-
-        try:
-            user_obj = Users.get_user_by_id(user_id)
-            model = body.get("model")
-
-            payload = {
-                "model": model,
-                "messages": [
-                    {
-                        "role": "system",
-                        "content": "You are a helpful assistant. Generate a short, concise title (max 10 words) for the following text. Do not use quotes. Only output the title.",
-                    },
-                    {"role": "user", "content": content[:2000]},  # Limit content length
-                ],
-                "stream": False,
-            }
-
-            response = await generate_chat_completion(request, payload, user_obj)
-            if response and "choices" in response:
-                return response["choices"][0]["message"]["content"].strip()
-        except Exception as e:
-            logger.error(f"Error generating title: {e}")
-
-        return ""
-
-    def extract_title(self, content: str) -> str:
-        """从 Markdown 内容提取一级/二级标题"""
-        lines = content.split("\n")
-        for line in lines:
-            # 仅匹配 h1-h2 标题
-            match = re.match(r"^#{1,2}\s+(.+)$", line.strip())
-            if match:
-                return match.group(1).strip()
-        return ""
-
-    def extract_chat_title(self, body: dict) -> str:
-        """从请求体中提取会话标题"""
-        if not isinstance(body, dict):
-            return ""
-
-        candidates = []
-
-        for key in ("chat", "conversation"):
-            if isinstance(body.get(key), dict):
-                candidates.append(body.get(key, {}).get("title", ""))
-
-        for key in ("title", "chat_title"):
-            value = body.get(key)
-            if isinstance(value, str):
-                candidates.append(value)
-
-        for candidate in candidates:
-            if candidate and isinstance(candidate, str):
-                return candidate.strip()
-        return ""
-
-    def extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
-        """从 body 或 metadata 中提取 chat_id"""
-        if isinstance(body, dict):
-            chat_id = body.get("chat_id") or body.get("id")
-            if isinstance(chat_id, str) and chat_id.strip():
-                return chat_id.strip()
-
-            for key in ("chat", "conversation"):
-                nested = body.get(key)
-                if isinstance(nested, dict):
-                    nested_id = nested.get("id") or nested.get("chat_id")
-                    if isinstance(nested_id, str) and nested_id.strip():
-                        return nested_id.strip()
-        if isinstance(metadata, dict):
-            chat_id = metadata.get("chat_id")
-            if isinstance(chat_id, str) and chat_id.strip():
-                return chat_id.strip()
-        return ""
-
-    async def fetch_chat_title(self, chat_id: str, user_id: str = "") -> str:
-        """根据 chat_id 从数据库获取标题"""
-        if not chat_id:
-            return ""
-
-        def _load_chat():
-            if user_id:
-                return Chats.get_chat_by_id_and_user_id(id=chat_id, user_id=user_id)
-            return Chats.get_chat_by_id(chat_id)
-
-        try:
-            chat = await asyncio.to_thread(_load_chat)
-        except Exception as exc:
-            logger.warning(f"加载聊天 {chat_id} 失败: {exc}")
-            return ""
-
-        if not chat:
-            return ""
-
-        data = getattr(chat, "chat", {}) or {}
-        title = data.get("title") or getattr(chat, "title", "")
-        return title.strip() if isinstance(title, str) else ""
-
-    def clean_filename(self, name: str) -> str:
-        """清理文件名中的非法字符"""
-        return re.sub(r'[\\/*?:"<>|]', "", name).strip()[:50]
-
-    def markdown_to_docx(
-        self, markdown_text: str, top_heading: str = "", has_h1: bool = False
-    ) -> Document:
-        """
-        将 Markdown 文本转换为 Word 文档
-        支持：标题、段落、粗体、斜体、代码块、列表、表格、链接
-        """
-        doc = Document()
-
-        # 设置默认中文字体
-        self.set_document_default_font(doc)
-
-        # 若正文无一级标题且有对话标题，则作为一级标题写入
-        if top_heading and not has_h1:
-            self.add_heading(doc, top_heading, 1)
-
-        lines = markdown_text.split("\n")
-        i = 0
-        in_code_block = False
-        code_block_content = []
-        code_block_lang = ""
-        in_list = False
-        list_items = []
-        list_type = None  # 'ordered' or 'unordered'
-
-        while i < len(lines):
-            line = lines[i]
-
-            # 处理代码块
-            if line.strip().startswith("```"):
-                if not in_code_block:
-                    # 先处理之前积累的列表
-                    if in_list and list_items:
-                        self.add_list_to_doc(doc, list_items, list_type)
-                        list_items = []
-                        in_list = False
-
-                    in_code_block = True
-                    code_block_lang = line.strip()[3:].strip()
-                    code_block_content = []
-                else:
-                    # 代码块结束
-                    in_code_block = False
-                    self.add_code_block(
-                        doc, "\n".join(code_block_content), code_block_lang
-                    )
-                    code_block_content = []
-                    code_block_lang = ""
-                i += 1
-                continue
-
-            if in_code_block:
-                code_block_content.append(line)
-                i += 1
-                continue
-
-            # 处理表格
-            if line.strip().startswith("|") and line.strip().endswith("|"):
-                # 先处理之前积累的列表
-                if in_list and list_items:
-                    self.add_list_to_doc(doc, list_items, list_type)
-                    list_items = []
-                    in_list = False
-
-                table_lines = []
-                while i < len(lines) and lines[i].strip().startswith("|"):
-                    table_lines.append(lines[i])
-                    i += 1
-                self.add_table(doc, table_lines)
-                continue
-
-            # 处理标题
-            header_match = re.match(r"^(#{1,6})\s+(.+)$", line.strip())
-            if header_match:
-                # 先处理之前积累的列表
-                if in_list and list_items:
-                    self.add_list_to_doc(doc, list_items, list_type)
-                    list_items = []
-                    in_list = False
-
-                level = len(header_match.group(1))
-                text = header_match.group(2)
-                self.add_heading(doc, text, level)
-                i += 1
-                continue
-
-            # 处理无序列表
-            unordered_match = re.match(r"^(\s*)[-*+]\s+(.+)$", line)
-            if unordered_match:
-                if not in_list or list_type != "unordered":
-                    if in_list and list_items:
-                        self.add_list_to_doc(doc, list_items, list_type)
-                        list_items = []
-                    in_list = True
-                    list_type = "unordered"
-                indent = len(unordered_match.group(1)) // 2
-                list_items.append((indent, unordered_match.group(2)))
-                i += 1
-                continue
-
-            # 处理有序列表
-            ordered_match = re.match(r"^(\s*)\d+[.)]\s+(.+)$", line)
-            if ordered_match:
-                if not in_list or list_type != "ordered":
-                    if in_list and list_items:
-                        self.add_list_to_doc(doc, list_items, list_type)
-                        list_items = []
-                    in_list = True
-                    list_type = "ordered"
-                indent = len(ordered_match.group(1)) // 2
-                list_items.append((indent, ordered_match.group(2)))
-                i += 1
-                continue
-
-            # 处理引用块
-            if line.strip().startswith(">"):
-                # 先处理之前积累的列表
-                if in_list and list_items:
-                    self.add_list_to_doc(doc, list_items, list_type)
-                    list_items = []
-                    in_list = False
-
-                # 收集连续的引用行
-                blockquote_lines = []
-                while i < len(lines) and lines[i].strip().startswith(">"):
-                    # 移除开头的 > 和可能的空格
-                    quote_line = re.sub(r"^>\s?", "", lines[i])
-                    blockquote_lines.append(quote_line)
-                    i += 1
-                self.add_blockquote(doc, "\n".join(blockquote_lines))
-                continue
-
-            # 处理水平分割线
-            if re.match(r"^[-*_]{3,}$", line.strip()):
-                # 先处理之前积累的列表
-                if in_list and list_items:
-                    self.add_list_to_doc(doc, list_items, list_type)
-                    list_items = []
-                    in_list = False
-
-                self.add_horizontal_rule(doc)
-                i += 1
-                continue
-
-            # 处理空行
-            if not line.strip():
-                # 列表结束
-                if in_list and list_items:
-                    self.add_list_to_doc(doc, list_items, list_type)
-                    list_items = []
-                    in_list = False
-                i += 1
-                continue
-
-            # 处理普通段落
-            if in_list and list_items:
-                self.add_list_to_doc(doc, list_items, list_type)
-                list_items = []
-                in_list = False
-
-            self.add_paragraph(doc, line)
-            i += 1
-
-        # 处理剩余的列表
-        if in_list and list_items:
-            self.add_list_to_doc(doc, list_items, list_type)
-
-        return doc
-
-    def set_document_default_font(self, doc: Document):
-        """设置文档默认字体，确保中英文都正常显示"""
-        # 设置正文样式
-        style = doc.styles["Normal"]
-        font = style.font
-        font.name = "Times New Roman"  # 英文字体
-        font.size = Pt(11)
-
-        # 设置中文字体
-        style._element.rPr.rFonts.set(qn("w:eastAsia"), "宋体")
-
-        # 设置段落格式
-        paragraph_format = style.paragraph_format
-        paragraph_format.line_spacing_rule = WD_LINE_SPACING.ONE_POINT_FIVE
-        paragraph_format.space_after = Pt(6)
-
-    def add_heading(self, doc: Document, text: str, level: int):
-        """添加标题"""
-        # Word 标题级别从 0 开始，Markdown 从 1 开始
-        heading_level = min(level, 9)  # Word 最多支持 Heading 9
-        heading = doc.add_heading(level=heading_level)
-
-        # 解析并添加格式化文本
-        self.add_formatted_text(heading, text)
-
-        # 设置中文字体
-        for run in heading.runs:
-            run.font.name = "Times New Roman"
-            run._element.rPr.rFonts.set(qn("w:eastAsia"), "黑体")
-            run.font.color.rgb = RGBColor(0, 0, 0)
-
-    def add_paragraph(self, doc: Document, text: str):
-        """添加段落，支持内联格式"""
-        paragraph = doc.add_paragraph()
-        self.add_formatted_text(paragraph, text)
-
-        # 设置中文字体
-        for run in paragraph.runs:
-            run.font.name = "Times New Roman"
-            run._element.rPr.rFonts.set(qn("w:eastAsia"), "宋体")
-
-    def add_formatted_text(self, paragraph, text: str):
-        """
-        解析 Markdown 内联格式并添加到段落
-        支持：粗体、斜体、行内代码、链接、删除线
-        """
-        # 定义格式化模式
-        patterns = [
-            # 粗斜体 ***text*** 或 ___text___
-            (r"\*\*\*(.+?)\*\*\*|___(.+?)___", {"bold": True, "italic": True}),
-            # 粗体 **text** 或 __text__
-            (r"\*\*(.+?)\*\*|__(.+?)__", {"bold": True}),
-            # 斜体 *text* 或 _text_
-            (
-                r"(?<!\*)\*(?!\*)(.+?)(?<!\*)\*(?!\*)|(?<!_)_(?!_)(.+?)(?<!_)_(?!_)",
-                {"italic": True},
-            ),
-            # 行内代码 `code`
-            (r"`([^`]+)`", {"code": True}),
-            # 链接 [text](url)
-            (r"\[([^\]]+)\]\(([^)]+)\)", {"link": True}),
-            # 删除线 ~~text~~
-            (r"~~(.+?)~~", {"strike": True}),
-        ]
-
-        # 简化处理：逐段解析
-        remaining = text
-        last_end = 0
-
-        # 合并所有匹配项
-        all_matches = []
-
-        for pattern, style in patterns:
-            for match in re.finditer(pattern, text):
-                # 获取匹配的文本内容
-                groups = match.groups()
-                matched_text = next((g for g in groups if g is not None), "")
-                all_matches.append(
-                    {
-                        "start": match.start(),
-                        "end": match.end(),
-                        "text": matched_text,
-                        "style": style,
-                        "full_match": match.group(0),
-                        "url": (
-                            groups[1] if style.get("link") and len(groups) > 1 else None
-                        ),
-                    }
-                )
-
-        # 按位置排序
-        all_matches.sort(key=lambda x: x["start"])
-
-        # 移除重叠的匹配
-        filtered_matches = []
-        last_end = 0
-        for m in all_matches:
-            if m["start"] >= last_end:
-                filtered_matches.append(m)
-                last_end = m["end"]
-
-        # 构建最终文本
-        pos = 0
-        for match in filtered_matches:
-            # 添加匹配前的普通文本
-            if match["start"] > pos:
-                plain_text = text[pos : match["start"]]
-                if plain_text:
-                    paragraph.add_run(plain_text)
-
-            # 添加格式化文本
-            style = match["style"]
-            run_text = match["text"]
-
-            if style.get("link"):
-                # 链接处理
-                run = paragraph.add_run(run_text)
-                run.font.color.rgb = RGBColor(0, 0, 255)
-                run.font.underline = True
-            elif style.get("code"):
-                # 行内代码
-                run = paragraph.add_run(run_text)
-                run.font.name = "Consolas"
-                run._element.rPr.rFonts.set(qn("w:eastAsia"), "SimHei")
-                run.font.size = Pt(10)
-                # 添加背景色
-                shading = OxmlElement("w:shd")
-                shading.set(qn("w:fill"), "E8E8E8")
-                run._element.rPr.append(shading)
-            else:
-                run = paragraph.add_run(run_text)
-                if style.get("bold"):
-                    run.bold = True
-                if style.get("italic"):
-                    run.italic = True
-                if style.get("strike"):
-                    run.font.strike = True
-
-            pos = match["end"]
-
-        # 添加剩余的普通文本
-        if pos < len(text):
-            paragraph.add_run(text[pos:])
-
-    def add_code_block(self, doc: Document, code: str, language: str = ""):
-        """添加代码块，支持语法高亮"""
-        # 语法高亮颜色映射 (基于常见的 IDE 配色)
-        TOKEN_COLORS = {
-            Token.Keyword: RGBColor(0, 92, 197),  # macOS 风格蓝 - 关键字
-            Token.Keyword.Constant: RGBColor(0, 92, 197),
-            Token.Keyword.Declaration: RGBColor(0, 92, 197),
-            Token.Keyword.Namespace: RGBColor(0, 92, 197),
-            Token.Keyword.Type: RGBColor(0, 92, 197),
-            Token.Name.Function: RGBColor(0, 0, 0),  # 函数名保持黑色
-            Token.Name.Class: RGBColor(38, 82, 120),  # 深青蓝 - 类名
-            Token.Name.Decorator: RGBColor(170, 51, 0),  # 暖橙 - 装饰器
-            Token.Name.Builtin: RGBColor(0, 110, 71),  # 墨绿 - 内置
-            Token.String: RGBColor(196, 26, 22),  # 红色 - 字符串
-            Token.String.Doc: RGBColor(109, 120, 133),  # 灰 - 文档字符串
-            Token.Comment: RGBColor(109, 120, 133),  # 灰 - 注释
-            Token.Comment.Single: RGBColor(109, 120, 133),
-            Token.Comment.Multiline: RGBColor(109, 120, 133),
-            Token.Number: RGBColor(28, 0, 207),  # 靛蓝 - 数字
-            Token.Number.Integer: RGBColor(28, 0, 207),
-            Token.Number.Float: RGBColor(28, 0, 207),
-            Token.Operator: RGBColor(90, 99, 120),  # 灰蓝 - 运算符
-            Token.Punctuation: RGBColor(0, 0, 0),  # 黑色 - 标点
-        }
-
-        def get_token_color(token_type):
-            """递归查找 token 颜色"""
-            while token_type:
-                if token_type in TOKEN_COLORS:
-                    return TOKEN_COLORS[token_type]
-                token_type = token_type.parent
-            return None
-
-        # 添加语言标签（如果有）
-        if language:
-            lang_para = doc.add_paragraph()
-            lang_para.paragraph_format.space_before = Pt(6)
-            lang_para.paragraph_format.space_after = Pt(0)
-            lang_para.paragraph_format.left_indent = Cm(0.5)
-            lang_run = lang_para.add_run(language.upper())
-            lang_run.font.name = "Consolas"
-            lang_run.font.size = Pt(8)
-            lang_run.font.color.rgb = RGBColor(100, 100, 100)
-            lang_run.font.bold = True
-
-        # 添加代码块段落
-        paragraph = doc.add_paragraph()
-        paragraph.paragraph_format.left_indent = Cm(0.5)
-        paragraph.paragraph_format.space_before = Pt(3) if language else Pt(6)
-        paragraph.paragraph_format.space_after = Pt(6)
-
-        # 添加浅灰色背景
-        shading = OxmlElement("w:shd")
-        shading.set(qn("w:fill"), "F7F7F7")
-        paragraph._element.pPr.append(shading)
-
-        # 尝试使用 Pygments 进行语法高亮
-        if PYGMENTS_AVAILABLE and language:
-            try:
-                lexer = get_lexer_by_name(language, stripall=False)
-            except Exception:
-                lexer = TextLexer()
-
-            tokens = list(lex(code, lexer))
-
-            for token_type, token_value in tokens:
-                if not token_value:
-                    continue
-                run = paragraph.add_run(token_value)
-                run.font.name = "Consolas"
-                run._element.rPr.rFonts.set(qn("w:eastAsia"), "SimHei")
-                run.font.size = Pt(10)
-
-                # 应用颜色
-                color = get_token_color(token_type)
-                if color:
-                    run.font.color.rgb = color
-
-                # 关键字加粗
-                if token_type in Token.Keyword:
-                    run.font.bold = True
-        else:
-            # 无语法高亮，纯文本显示
-            run = paragraph.add_run(code)
-            run.font.name = "Consolas"
-            run._element.rPr.rFonts.set(qn("w:eastAsia"), "SimHei")
-            run.font.size = Pt(10)
-
-    def add_table(self, doc: Document, table_lines: List[str]):
-        """添加表格，支持表头底色与隔行底色"""
-        if len(table_lines) < 2:
-            return
-
-        def _set_cell_shading(cell, fill: str):
-            tc_pr = cell._element.get_or_add_tcPr()
-            shd = OxmlElement("w:shd")
-            shd.set(qn("w:fill"), fill)
-            tc_pr.append(shd)
-
-        header_fill = "F2F2F2"
-        zebra_fill = "FBFBFB"
-
-        # 解析表格数据
-        rows = []
-        for line in table_lines:
-            cells = [cell.strip() for cell in line.strip().strip("|").split("|")]
-            # 跳过分隔行
-            if all(re.fullmatch(r"[-:]+", cell) for cell in cells):
-                continue
-            rows.append(cells)
-
-        if not rows:
-            return
-
-        # 确定列数
-        num_cols = max(len(row) for row in rows)
-
-        # 创建表格
-        table = doc.add_table(rows=len(rows), cols=num_cols)
-        table.style = "Table Grid"
-        table.alignment = WD_TABLE_ALIGNMENT.CENTER
-
-        # 填充表格
-        for row_idx, row_data in enumerate(rows):
-            row = table.rows[row_idx]
-            for col_idx, cell_text in enumerate(row_data):
-                if col_idx < num_cols:
-                    cell = row.cells[col_idx]
-                    # 清除默认段落
-                    cell.paragraphs[0].clear()
-                    para = cell.paragraphs[0]
-                    para.paragraph_format.space_after = Pt(3)
-                    para.paragraph_format.space_before = Pt(1)
-                    para.alignment = WD_ALIGN_PARAGRAPH.LEFT
-
-                    self.add_formatted_text(para, cell_text)
-
-                    # 设置单元格字体
-                    for run in para.runs:
-                        run.font.name = "Times New Roman"
-                        run._element.rPr.rFonts.set(qn("w:eastAsia"), "宋体")
-                        run.font.size = Pt(10)
-
-                    # 表头加粗并填充底色
-                    if row_idx == 0:
-                        for run in para.runs:
-                            run.bold = True
-                        _set_cell_shading(cell, header_fill)
-                    # 隔行底色
-                    elif row_idx % 2 == 1:
-                        _set_cell_shading(cell, zebra_fill)
-
-        # 统一列对齐为左对齐，避免居中导致阅读困难
-        for row in table.rows:
-            for cell in row.cells:
-                for para in cell.paragraphs:
-                    para.alignment = WD_ALIGN_PARAGRAPH.LEFT
-
-    def add_list_to_doc(
-        self, doc: Document, items: List[Tuple[int, str]], list_type: str
-    ):
-        """添加列表"""
-        for indent, text in items:
-            paragraph = doc.add_paragraph()
-
-            if list_type == "unordered":
-                # 无序列表使用项目符号
-                paragraph.style = "List Bullet"
-            else:
-                # 有序列表使用编号
-                paragraph.style = "List Number"
-
-            # 设置缩进
-            paragraph.paragraph_format.left_indent = Cm(0.5 * (indent + 1))
-
-            # 添加格式化文本
-            self.add_formatted_text(paragraph, text)
-
-            # 设置字体
-            for run in paragraph.runs:
-                run.font.name = "Times New Roman"
-                run._element.rPr.rFonts.set(qn("w:eastAsia"), "宋体")
-
-    def add_horizontal_rule(self, doc: Document):
-        """添加水平分割线"""
-        paragraph = doc.add_paragraph()
-        paragraph.paragraph_format.space_before = Pt(12)
-        paragraph.paragraph_format.space_after = Pt(12)
-
-        # 添加底部边框作为分割线
-        pPr = paragraph._element.get_or_add_pPr()
-        pBdr = OxmlElement("w:pBdr")
-        bottom = OxmlElement("w:bottom")
-        bottom.set(qn("w:val"), "single")
-        bottom.set(qn("w:sz"), "6")
-        bottom.set(qn("w:space"), "1")
-        bottom.set(qn("w:color"), "auto")
-        pBdr.append(bottom)
-        pPr.append(pBdr)
-
-    def add_blockquote(self, doc: Document, text: str):
-        """添加引用块，带有左侧边框和灰色背景"""
-        for line in text.split("\n"):
-            paragraph = doc.add_paragraph()
-            paragraph.paragraph_format.left_indent = Cm(1.0)
-            paragraph.paragraph_format.space_before = Pt(3)
-            paragraph.paragraph_format.space_after = Pt(3)
-
-            # 添加左侧边框
-            pPr = paragraph._element.get_or_add_pPr()
-            pBdr = OxmlElement("w:pBdr")
-            left = OxmlElement("w:left")
-            left.set(qn("w:val"), "single")
-            left.set(qn("w:sz"), "24")  # 边框粗细
-            left.set(qn("w:space"), "4")  # 边框与文字间距
-            left.set(qn("w:color"), "CCCCCC")  # 灰色边框
-            pBdr.append(left)
-            pPr.append(pBdr)
-
-            # 添加浅灰色背景
-            shading = OxmlElement("w:shd")
-            shading.set(qn("w:fill"), "F9F9F9")
-            pPr.append(shading)
-
-            # 添加格式化文本
-            self.add_formatted_text(paragraph, line)
-
-            # 设置字体为斜体灰色
-            for run in paragraph.runs:
-                run.font.name = "Times New Roman"
-                run._element.rPr.rFonts.set(qn("w:eastAsia"), "楷体")
-                run.font.color.rgb = RGBColor(85, 85, 85)  # 深灰色文字
-                run.italic = True

plugins/actions/export_to_excel/README.md

@@ -2,12 +2,33 @@
 
 This plugin allows you to export your chat history to an Excel (.xlsx) file directly from the chat interface.
 
+## What's New in v0.3.6
+
+- **OpenWebUI-Style Theme**: Modern dark header (#1f2937) with light gray zebra striping for better readability.
+- **Zebra Striping**: Alternating row colors (#ffffff / #f3f4f6) for improved visual scanning.
+- **Smart Data Type Conversion**: Automatically converts columns to numeric or datetime types with fallback to string.
+- **Full Cell Bold/Italic**: Supports full cell Markdown bold (`**text**`) and italic (`*text*`) formatting in Excel.
+- **Partial Markdown Cleanup**: Automatically removes partial Markdown formatting symbols (e.g., `Some **bold** text` → `Some bold text`) for cleaner Excel output.
+- **Export Scope**: Added `EXPORT_SCOPE` valve to choose between exporting tables from the "Last Message" (default) or "All Messages".
+- **Smart Sheet Naming**: Automatically names sheets based on Markdown headers, AI titles (if enabled), or message index (e.g., `Msg1-Tab1`).
+- **Multiple Tables Support**: Improved handling of multiple tables within single or multiple messages.
+- **Smart Filename Generation**: Supports generating filenames based on Chat Title, AI Summary, or Markdown Headers.
+- **Configuration Options**: Added `TITLE_SOURCE` setting to control filename generation strategy.
+- **AI Title Generation**: Added `MODEL_ID` setting to specify the model for AI title generation, with progress notifications.
+
 ## Features
 
 - **One-Click Export**: Adds an "Export to Excel" button to the chat.
 - **Automatic Header Extraction**: Intelligently identifies table headers from the chat content.
 - **Multi-Table Support**: Handles multiple tables within a single chat session.
 
+## Configuration
+
+- **Title Source**: Choose how the filename is generated:
+  - `chat_title`: Use the chat title (default).
+  - `ai_generated`: Use AI to generate a concise title from the content.
+  - `markdown_title`: Extract the first H1/H2 header from the markdown content.
+
 ## Usage
 
 1. Install the plugin.

plugins/actions/export_to_excel/README_CN.md

@@ -2,16 +2,37 @@
 
 此插件允许你直接从聊天界面将对话历史导出为 Excel (.xlsx) 文件。
 
+## v0.3.6 更新内容
+
+- **OpenWebUI 风格主题**：现代深灰表头 (#1f2937)，搭配浅灰斑马纹，提升可读性。
+- **斑马纹效果**：隔行变色（#ffffff / #f3f4f6），方便视觉扫描。
+- **智能数据类型转换**：自动将列转换为数字或日期类型，无法转换时保持字符串。
+- **全单元格粗体/斜体**：支持 Excel 中的全单元格 Markdown 粗体 (`**text**`) 和斜体 (`*text*`) 格式。
+- **部分 Markdown 清理**：自动移除部分 Markdown 格式符号（如 `部分**加粗**文本` → `部分加粗文本`），使 Excel 输出更整洁。
+- **导出范围**: 新增 `EXPORT_SCOPE` 配置项，可选择导出"最后一条消息"（默认）或"所有消息"中的表格。
+- **智能 Sheet 命名**: 根据 Markdown 标题、AI 标题（如启用）或消息索引（如 `消息1-表1`）自动命名 Sheet。
+- **多表格支持**: 优化了对单条或多条消息中包含多个表格的处理。
+- **智能文件名生成**：支持根据对话标题、AI 总结或 Markdown 标题生成文件名。
+- **配置选项**：新增 `TITLE_SOURCE` 设置，用于控制文件名生成策略。
+- **AI 标题生成**：新增 `MODEL_ID` 设置用于指定 AI 标题生成模型，并支持生成进度通知。
+
 ## 功能特点
 
-- **一键导出**：在聊天界面添加“导出为 Excel”按钮。
+- **一键导出**：在聊天界面添加"导出为 Excel"按钮。
 - **自动表头提取**：智能识别聊天内容中的表格标题。
 - **多表支持**：支持处理单次对话中的多个表格。
 
+## 配置
+
+- **标题来源 (Title Source)**：选择文件名的生成方式：
+  - `chat_title`：使用对话标题（默认）。
+  - `ai_generated`：使用 AI 根据内容生成简洁标题。
+  - `markdown_title`：提取 Markdown 内容中的第一个 H1/H2 标题。
+
 ## 使用方法
 
 1. 安装插件。
-2. 在任意对话中，点击“导出为 Excel”按钮。
+2. 在任意对话中，点击"导出为 Excel"按钮。
 3. 文件将自动下载到你的设备。
 
 ## 作者

plugins/actions/export_to_excel/export_to_excel.py

@@ -3,9 +3,10 @@ title: Export to Excel
 author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.3.3
+version: 0.3.7
+openwebui_id: 244b8f9d-7459-47d6-84d3-c7ae8e3ec710
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxwYXRoIGQ9Ik0xNSAySDZhMiAyIDAgMCAwLTIgMnYxNmEyIDIgMCAwIDAgMiAyaDEyYTIgMiAwIDAgMCAyLTJWN1oiLz48cGF0aCBkPSJNMTQgMnY0YTIgMiAwIDAgMCAyIDJoNCIvPjxwYXRoIGQ9Ik04IDEzaDIiLz48cGF0aCBkPSJNMTQgMTNoMiIvPjxwYXRoIGQ9Ik04IDE3aDIiLz48cGF0aCBkPSJNMTQgMTdoMiIvPjwvc3ZnPg==
-description: Exports the current chat history to an Excel (.xlsx) file, with automatic header extraction.
+description: Extracts tables from chat messages and exports them to Excel (.xlsx) files with smart formatting.
 """
 
 import os
@@ -15,14 +16,33 @@ import base64
 from fastapi import FastAPI, HTTPException
 from typing import Optional, Callable, Awaitable, Any, List, Dict
 import datetime
+import asyncio
+from open_webui.models.chats import Chats
+from open_webui.models.users import Users
+from open_webui.utils.chat import generate_chat_completion
+from pydantic import BaseModel, Field
+from typing import Literal
 
 app = FastAPI()
 
 
 class Action:
+    class Valves(BaseModel):
+        TITLE_SOURCE: Literal["chat_title", "ai_generated", "markdown_title"] = Field(
+            default="chat_title",
+            description="Title Source: 'chat_title' (Chat Title), 'ai_generated' (AI Generated), 'markdown_title' (Markdown Title)",
+        )
+        EXPORT_SCOPE: Literal["last_message", "all_messages"] = Field(
+            default="last_message",
+            description="Export Scope: 'last_message' (Last Message Only), 'all_messages' (All Messages)",
+        )
+        MODEL_ID: str = Field(
+            default="",
+            description="Model ID for AI title generation. Leave empty to use the current chat model.",
+        )
 
     def __init__(self):
-        pass
+        self.valves = self.Valves()
 
     async def _send_notification(self, emitter: Callable, type: str, content: str):
         await emitter(
@@ -35,6 +55,7 @@ class Action:
         __user__=None,
         __event_emitter__=None,
         __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __request__: Optional[Any] = None,
     ):
         print(f"action:{__name__}")
         if isinstance(__user__, (list, tuple)):
@@ -53,8 +74,6 @@ class Action:
             user_id = __user__.get("id", "unknown_user")
 
         if __event_emitter__:
-            last_assistant_message = body["messages"][-1]
-
             await __event_emitter__(
                 {
                     "type": "status",
@@ -63,24 +82,180 @@ class Action:
             )
 
             try:
-                message_content = last_assistant_message["content"]
-                tables = self.extract_tables_from_message(message_content)
+                messages = body.get("messages", [])
+                if not messages:
+                    raise HTTPException(status_code=400, detail="No messages found.")
+
+                # Determine messages to process based on scope
+                target_messages = []
+                if self.valves.EXPORT_SCOPE == "all_messages":
+                    target_messages = messages
+                else:
+                    target_messages = [messages[-1]]
+
+                all_tables = []
+                all_sheet_names = []
+
+                # Process messages
+                for msg_index, msg in enumerate(target_messages):
+                    content = msg.get("content", "")
+                    tables = self.extract_tables_from_message(content)
 
                     if not tables:
-                    raise HTTPException(status_code=400, detail="No tables found.")
+                        continue
 
-                # Get dynamic filename and sheet names
-                workbook_name, sheet_names = self.generate_names_from_content(
-                    message_content, tables
+                    # Generate sheet names for this message's tables
+                    # If multiple messages, we need to ensure uniqueness across the whole workbook
+                    # We'll generate base names here and deduplicate later if needed,
+                    # or better: generate unique names on the fly.
+
+                    # Extract headers for this message
+                    headers = []
+                    lines = content.split("\n")
+                    for i, line in enumerate(lines):
+                        if re.match(r"^#{1,6}\s+", line):
+                            headers.append(
+                                {
+                                    "text": re.sub(r"^#{1,6}\s+", "", line).strip(),
+                                    "line_num": i,
+                                }
                             )
 
-                # Use optimized filename generation logic
+                    for table_index, table in enumerate(tables):
+                        sheet_name = ""
+
+                        # 1. Try Markdown Header (closest above)
+                        table_start_line = table["start_line"] - 1
+                        closest_header_text = None
+                        candidate_headers = [
+                            h for h in headers if h["line_num"] < table_start_line
+                        ]
+                        if candidate_headers:
+                            closest_header = max(
+                                candidate_headers, key=lambda x: x["line_num"]
+                            )
+                            closest_header_text = closest_header["text"]
+
+                        if closest_header_text:
+                            sheet_name = self.clean_sheet_name(closest_header_text)
+
+                        # 2. AI Generated (Only if explicitly enabled and we have a request object)
+                        # Note: Generating titles for EVERY table in all messages might be too slow/expensive.
+                        # We'll skip this for 'all_messages' scope to avoid timeout, unless it's just one message.
+                        if (
+                            not sheet_name
+                            and self.valves.TITLE_SOURCE == "ai_generated"
+                            and len(target_messages) == 1
+                        ):
+                            # Logic for AI generation (simplified for now, reusing existing flow if possible)
+                            pass
+
+                        # 3. Fallback: Message Index
+                        if not sheet_name:
+                            if len(target_messages) > 1:
+                                # Use global message index (from original list if possible, but here we iterate target_messages)
+                                # Let's use the loop index.
+                                # If multiple tables in one message: "Msg 1 - Table 1"
+                                if len(tables) > 1:
+                                    sheet_name = f"Msg{msg_index+1}-Tab{table_index+1}"
+                                else:
+                                    sheet_name = f"Msg{msg_index+1}"
+                            else:
+                                # Single message (last_message scope)
+                                if len(tables) > 1:
+                                    sheet_name = f"Table {table_index+1}"
+                                else:
+                                    sheet_name = "Sheet1"
+
+                        all_tables.append(table)
+                        all_sheet_names.append(sheet_name)
+
+                if not all_tables:
+                    raise HTTPException(
+                        status_code=400, detail="No tables found in the selected scope."
+                    )
+
+                # Deduplicate sheet names
+                final_sheet_names = []
+                seen_names = {}
+                for name in all_sheet_names:
+                    base_name = name
+                    counter = 1
+                    while name in seen_names:
+                        name = f"{base_name} ({counter})"
+                        counter += 1
+                    seen_names[name] = True
+                    final_sheet_names.append(name)
+
+                # Notify user about the number of tables found
+                table_count = len(all_tables)
+                if self.valves.EXPORT_SCOPE == "all_messages":
+                    await self._send_notification(
+                        __event_emitter__,
+                        "info",
+                        f"Found {table_count} table(s) in all messages.",
+                    )
+                    # Wait a moment for user to see the notification before download dialog
+                    await asyncio.sleep(1.5)
+                # Generate Workbook Title (Filename)
+                # Use the title of the chat, or the first header of the first message with tables
+                title = ""
+                chat_id = self.extract_chat_id(body, None)
+                chat_title = ""
+                if chat_id:
+                    chat_title = await self.fetch_chat_title(chat_id, user_id)
+
+                if (
+                    self.valves.TITLE_SOURCE == "chat_title"
+                    or not self.valves.TITLE_SOURCE
+                ):
+                    title = chat_title
+                elif self.valves.TITLE_SOURCE == "ai_generated":
+                    # Use AI to generate a title based on message content
+                    if target_messages and __request__:
+                        # Get content from the first message with tables
+                        content_for_title = ""
+                        for msg in target_messages:
+                            msg_content = msg.get("content", "")
+                            if msg_content:
+                                content_for_title = msg_content
+                                break
+                        if content_for_title:
+                            title = await self.generate_title_using_ai(
+                                body,
+                                content_for_title,
+                                user_id,
+                                __request__,
+                                __event_emitter__,
+                            )
+                elif self.valves.TITLE_SOURCE == "markdown_title":
+                    # Try to find first header in the first message that has content
+                    for msg in target_messages:
+                        extracted = self.extract_title(msg.get("content", ""))
+                        if extracted:
+                            title = extracted
+                            break
+
+                # Fallback for filename
+                if not title:
+                    if chat_title:
+                        title = chat_title
+                    else:
+                        # Try extracting from content again if not already tried
+                        if self.valves.TITLE_SOURCE != "markdown_title":
+                            for msg in target_messages:
+                                extracted = self.extract_title(msg.get("content", ""))
+                                if extracted:
+                                    title = extracted
+                                    break
+
                 current_datetime = datetime.datetime.now()
                 formatted_date = current_datetime.strftime("%Y%m%d")
 
-                # If no title found, use user_yyyymmdd format
-                if not workbook_name:
+                if not title:
                     workbook_name = f"{user_name}_{formatted_date}"
+                else:
+                    workbook_name = self.clean_filename(title)
 
                 filename = f"{workbook_name}.xlsx"
                 excel_file_path = os.path.join(
@@ -89,8 +264,10 @@ class Action:
 
                 os.makedirs(os.path.dirname(excel_file_path), exist_ok=True)
 
-                # Save tables to Excel (using enhanced formatting)
-                self.save_tables_to_excel_enhanced(tables, excel_file_path, sheet_names)
+                # Save tables to Excel
+                self.save_tables_to_excel_enhanced(
+                    all_tables, excel_file_path, final_sheet_names
+                )
 
                 # Trigger file download
                 if __event_call__:
@@ -172,6 +349,149 @@ class Action:
                     __event_emitter__, "error", "No tables found to export!"
                 )
 
+    async def generate_title_using_ai(
+        self,
+        body: dict,
+        content: str,
+        user_id: str,
+        request: Any,
+        event_emitter: Callable = None,
+    ) -> str:
+        if not request:
+            return ""
+
+        try:
+            user_obj = Users.get_user_by_id(user_id)
+            # Use configured MODEL_ID or fallback to current chat model
+            model = (
+                self.valves.MODEL_ID.strip()
+                if self.valves.MODEL_ID
+                else body.get("model")
+            )
+
+            payload = {
+                "model": model,
+                "messages": [
+                    {
+                        "role": "system",
+                        "content": "You are a helpful assistant. Generate a short, concise filename (max 10 words) for an Excel export based on the following content. Do not use quotes or file extensions. Avoid special characters that are invalid in filenames. Only output the filename.",
+                    },
+                    {"role": "user", "content": content[:2000]},  # Limit content length
+                ],
+                "stream": False,
+            }
+
+            # Define the generation task
+            async def generate_task():
+                return await generate_chat_completion(request, payload, user_obj)
+
+            # Define the notification task
+            async def notification_task():
+                # Send initial notification immediately
+                if event_emitter:
+                    await self._send_notification(
+                        event_emitter,
+                        "info",
+                        "AI is generating a filename for your Excel file...",
+                    )
+
+                # Subsequent notifications every 5 seconds
+                while True:
+                    await asyncio.sleep(5)
+                    if event_emitter:
+                        await self._send_notification(
+                            event_emitter,
+                            "info",
+                            "Still generating filename, please be patient...",
+                        )
+
+            # Run tasks concurrently
+            gen_future = asyncio.ensure_future(generate_task())
+            notify_future = asyncio.ensure_future(notification_task())
+
+            done, pending = await asyncio.wait(
+                [gen_future, notify_future], return_when=asyncio.FIRST_COMPLETED
+            )
+
+            # Cancel notification task if generation is done
+            if not notify_future.done():
+                notify_future.cancel()
+
+            # Get result
+            if gen_future in done:
+                response = gen_future.result()
+                if response and "choices" in response:
+                    return response["choices"][0]["message"]["content"].strip()
+            else:
+                # Should not happen if return_when=FIRST_COMPLETED and we cancel notify
+                await gen_future
+                response = gen_future.result()
+                if response and "choices" in response:
+                    return response["choices"][0]["message"]["content"].strip()
+
+        except Exception as e:
+            print(f"Error generating title: {e}")
+            if event_emitter:
+                await self._send_notification(
+                    event_emitter,
+                    "warning",
+                    f"AI title generation failed, using default title. Error: {str(e)}",
+                )
+
+        return ""
+
+    def extract_title(self, content: str) -> str:
+        """Extract title from Markdown h1/h2 only"""
+        lines = content.split("\n")
+        for line in lines:
+            # Match h1-h2 headings only
+            match = re.match(r"^#{1,2}\s+(.+)$", line.strip())
+            if match:
+                return match.group(1).strip()
+        return ""
+
+    def extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract chat_id from body or metadata"""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id") or body.get("id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+            for key in ("chat", "conversation"):
+                nested = body.get(key)
+                if isinstance(nested, dict):
+                    nested_id = nested.get("id") or nested.get("chat_id")
+                    if isinstance(nested_id, str) and nested_id.strip():
+                        return nested_id.strip()
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+        return ""
+
+    async def fetch_chat_title(self, chat_id: str, user_id: str = "") -> str:
+        """Fetch chat title from database by chat_id"""
+        if not chat_id:
+            return ""
+
+        def _load_chat():
+            if user_id:
+                return Chats.get_chat_by_id_and_user_id(id=chat_id, user_id=user_id)
+            return Chats.get_chat_by_id(chat_id)
+
+        try:
+            chat = await asyncio.to_thread(_load_chat)
+        except Exception as exc:
+            print(f"Failed to load chat {chat_id}: {exc}")
+            return ""
+
+        if not chat:
+            return ""
+
+        data = getattr(chat, "chat", {}) or {}
+        title = data.get("title") or getattr(chat, "title", "")
+        return title.strip() if isinstance(title, str) else ""
+
     def extract_tables_from_message(self, message: str) -> List[Dict]:
         """
         Extract Markdown tables and their positions from message text
@@ -456,24 +776,51 @@ class Action:
             with pd.ExcelWriter(file_path, engine="xlsxwriter") as writer:
                 workbook = writer.book
 
+                # OpenWebUI-style theme colors
+                HEADER_BG = "#1f2937"  # Dark gray (matches OpenWebUI sidebar)
+                HEADER_FG = "#ffffff"  # White text
+                ROW_ODD_BG = "#ffffff"  # White for odd rows
+                ROW_EVEN_BG = "#f3f4f6"  # Light gray for even rows (zebra striping)
+                BORDER_COLOR = "#e5e7eb"  # Light border
+
                 # Define header style - Center aligned
                 header_format = workbook.add_format(
                     {
                         "bold": True,
-                        "font_size": 12,
-                        "font_color": "white",
-                        "bg_color": "#00abbd",
+                        "font_size": 11,
+                        "font_name": "Arial",
+                        "font_color": HEADER_FG,
+                        "bg_color": HEADER_BG,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
                         "align": "center",
                         "valign": "vcenter",
                         "text_wrap": True,
                     }
                 )
 
-                # Text cell style - Left aligned
+                # Text cell style - Left aligned (odd rows)
                 text_format = workbook.add_format(
                     {
+                        "font_name": "Arial",
+                        "font_size": 10,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                    }
+                )
+
+                # Text cell style - Left aligned (even rows - zebra)
+                text_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
                         "align": "left",
                         "valign": "vcenter",
                         "text_wrap": True,
@@ -482,14 +829,51 @@ class Action:
 
                 # Number cell style - Right aligned
                 number_format = workbook.add_format(
-                    {"border": 1, "align": "right", "valign": "vcenter"}
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "right",
+                        "valign": "vcenter",
+                    }
+                )
+
+                number_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
+                        "align": "right",
+                        "valign": "vcenter",
+                    }
                 )
 
                 # Integer format - Right aligned
                 integer_format = workbook.add_format(
                     {
                         "num_format": "0",
+                        "font_name": "Arial",
+                        "font_size": 10,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "right",
+                        "valign": "vcenter",
+                    }
+                )
+
+                integer_format_alt = workbook.add_format(
+                    {
+                        "num_format": "0",
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
                         "align": "right",
                         "valign": "vcenter",
                     }
@@ -499,7 +883,24 @@ class Action:
                 decimal_format = workbook.add_format(
                     {
                         "num_format": "0.00",
+                        "font_name": "Arial",
+                        "font_size": 10,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "right",
+                        "valign": "vcenter",
+                    }
+                )
+
+                decimal_format_alt = workbook.add_format(
+                    {
+                        "num_format": "0.00",
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
                         "align": "right",
                         "valign": "vcenter",
                     }
@@ -508,7 +909,24 @@ class Action:
                 # Date format - Center aligned
                 date_format = workbook.add_format(
                     {
+                        "font_name": "Arial",
+                        "font_size": 10,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "center",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                    }
+                )
+
+                date_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
                         "align": "center",
                         "valign": "vcenter",
                         "text_wrap": True,
@@ -518,12 +936,114 @@ class Action:
                 # Sequence format - Center aligned
                 sequence_format = workbook.add_format(
                     {
+                        "font_name": "Arial",
+                        "font_size": 10,
                         "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
                         "align": "center",
                         "valign": "vcenter",
                     }
                 )
 
+                sequence_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
+                        "align": "center",
+                        "valign": "vcenter",
+                    }
+                )
+
+                # Bold cell style (for full cell bolding)
+                text_bold_format = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                        "bold": True,
+                    }
+                )
+
+                text_bold_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                        "bold": True,
+                    }
+                )
+
+                # Italic cell style (for full cell italics)
+                text_italic_format = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_ODD_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                        "italic": True,
+                    }
+                )
+
+                text_italic_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Arial",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": ROW_EVEN_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                        "italic": True,
+                    }
+                )
+
+                # Code cell style (for inline code with highlight background)
+                CODE_BG = "#f0f0f0"  # Light gray background for code
+                text_code_format = workbook.add_format(
+                    {
+                        "font_name": "Consolas",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": CODE_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                    }
+                )
+
+                text_code_format_alt = workbook.add_format(
+                    {
+                        "font_name": "Consolas",
+                        "font_size": 10,
+                        "border": 1,
+                        "border_color": BORDER_COLOR,
+                        "bg_color": CODE_BG,
+                        "align": "left",
+                        "valign": "vcenter",
+                        "text_wrap": True,
+                    }
+                )
+
                 for i, table in enumerate(tables):
                     try:
                         table_data = table["data"]
@@ -565,12 +1085,18 @@ class Action:
 
                         print(f"DataFrame created with columns: {list(df.columns)}")
 
-                        # Fix pandas FutureWarning
+                        # Smart data type conversion using pandas infer_objects
                         for col in df.columns:
+                            # Try numeric conversion first
                             try:
                                 df[col] = pd.to_numeric(df[col])
                             except (ValueError, TypeError):
-                                pass
+                                # Try datetime conversion
+                                try:
+                                    df[col] = pd.to_datetime(df[col], errors="raise")
+                                except (ValueError, TypeError):
+                                    # Keep as string, use infer_objects for optimization
+                                    df[col] = df[col].infer_objects()
 
                         # Write data first (without header)
                         df.to_excel(
@@ -582,19 +1108,25 @@ class Action:
                         )
                         worksheet = writer.sheets[sheet_name]
 
-                        # Apply enhanced formatting
+                        # Apply enhanced formatting with zebra striping
+                        formats = {
+                            "header": header_format,
+                            "text": [text_format, text_format_alt],
+                            "number": [number_format, number_format_alt],
+                            "integer": [integer_format, integer_format_alt],
+                            "decimal": [decimal_format, decimal_format_alt],
+                            "date": [date_format, date_format_alt],
+                            "sequence": [sequence_format, sequence_format_alt],
+                            "bold": [text_bold_format, text_bold_format_alt],
+                            "italic": [text_italic_format, text_italic_format_alt],
+                            "code": [text_code_format, text_code_format_alt],
+                        }
                         self.apply_enhanced_formatting(
                             worksheet,
                             df,
                             headers,
                             workbook,
-                            header_format,
-                            text_format,
-                            number_format,
-                            integer_format,
-                            decimal_format,
-                            date_format,
-                            sequence_format,
+                            formats,
                         )
 
                     except Exception as e:
@@ -611,23 +1143,22 @@ class Action:
         df,
         headers,
         workbook,
-        header_format,
-        text_format,
-        number_format,
-        integer_format,
-        decimal_format,
-        date_format,
-        sequence_format,
+        formats,
     ):
         """
-        Apply enhanced formatting
-        - Header: Center aligned
+        Apply enhanced formatting with zebra striping
+        - Header: Center aligned (dark background)
         - Number: Right aligned
         - Text: Left aligned
         - Date: Center aligned
         - Sequence: Center aligned
+        - Zebra striping: alternating row colors
+        - Supports full cell Markdown bold (**text**) and italic (*text*)
         """
         try:
+            # Extract format from formats dict
+            header_format = formats["header"]
+
             # 1. Write headers (Center aligned)
             print(f"Writing headers with enhanced alignment: {headers}")
             for col_idx, header in enumerate(headers):
@@ -651,42 +1182,98 @@ class Action:
                 else:
                     column_types[col_idx] = "text"
 
-            # 3. Write and format data
+            # 3. Write and format data with zebra striping
             for row_idx, row in df.iterrows():
+                # Determine if odd or even row (0-indexed, so row 0 is odd visually as row 1)
+                is_alt_row = (
+                    row_idx % 2 == 1
+                )  # Even index = odd visual row, use alt format
+
                 for col_idx, value in enumerate(row):
                     content_type = column_types.get(col_idx, "text")
 
-                    # Select format based on content type
+                    # Select format based on content type and zebra striping
+                    fmt_idx = 1 if is_alt_row else 0
+
                     if content_type == "number":
                         # Number - Right aligned
                         if pd.api.types.is_numeric_dtype(df.iloc[:, col_idx]):
                             if pd.api.types.is_integer_dtype(df.iloc[:, col_idx]):
-                                current_format = integer_format
+                                current_format = formats["integer"][fmt_idx]
                             else:
                                 try:
                                     numeric_value = float(value)
                                     if numeric_value.is_integer():
-                                        current_format = integer_format
+                                        current_format = formats["integer"][fmt_idx]
                                         value = int(numeric_value)
                                     else:
-                                        current_format = decimal_format
+                                        current_format = formats["decimal"][fmt_idx]
                                 except (ValueError, TypeError):
-                                    current_format = decimal_format
+                                    current_format = formats["decimal"][fmt_idx]
                         else:
-                            current_format = number_format
+                            current_format = formats["number"][fmt_idx]
 
                     elif content_type == "date":
                         # Date - Center aligned
-                        current_format = date_format
+                        current_format = formats["date"][fmt_idx]
 
                     elif content_type == "sequence":
                         # Sequence - Center aligned
-                        current_format = sequence_format
+                        current_format = formats["sequence"][fmt_idx]
 
                     else:
                         # Text - Left aligned
-                        current_format = text_format
+                        current_format = formats["text"][fmt_idx]
 
+                    if content_type == "text" and isinstance(value, str):
+                        # Check for full cell bold (**text**)
+                        match_bold = re.fullmatch(r"\*\*(.+)\*\*", value.strip())
+                        # Check for full cell italic (*text*)
+                        match_italic = re.fullmatch(r"\*(.+)\*", value.strip())
+                        # Check for full cell code (`text`)
+                        match_code = re.fullmatch(r"`(.+)`", value.strip())
+
+                        if match_bold:
+                            # Extract content and apply bold format
+                            clean_value = match_bold.group(1)
+                            worksheet.write(
+                                row_idx + 1,
+                                col_idx,
+                                clean_value,
+                                formats["bold"][fmt_idx],
+                            )
+                        elif match_italic:
+                            # Extract content and apply italic format
+                            clean_value = match_italic.group(1)
+                            worksheet.write(
+                                row_idx + 1,
+                                col_idx,
+                                clean_value,
+                                formats["italic"][fmt_idx],
+                            )
+                        elif match_code:
+                            # Extract content and apply code format (highlighted)
+                            clean_value = match_code.group(1)
+                            worksheet.write(
+                                row_idx + 1,
+                                col_idx,
+                                clean_value,
+                                formats["code"][fmt_idx],
+                            )
+                        else:
+                            # Remove partial markdown formatting symbols (can't render partial formatting in Excel)
+                            # Remove bold markers **text** -> text
+                            clean_value = re.sub(r"\*\*(.+?)\*\*", r"\1", value)
+                            # Remove italic markers *text* -> text (but not inside **)
+                            clean_value = re.sub(
+                                r"(?<!\*)\*([^*]+)\*(?!\*)", r"\1", clean_value
+                            )
+                            # Remove code markers `text` -> text
+                            clean_value = re.sub(r"`(.+?)`", r"\1", clean_value)
+                            worksheet.write(
+                                row_idx + 1, col_idx, clean_value, current_format
+                            )
+                    else:
                         worksheet.write(row_idx + 1, col_idx, value, current_format)
 
             # 4. Auto-adjust column width
@@ -777,3 +1364,6 @@ class Action:
 
         except Exception as e:
             print(f"Error in basic formatting: {str(e)}")
+
+        except Exception as e:
+            print(f"Error in basic formatting: {str(e)}")

plugins/actions/export_to_excel/导出为Excel.py

@@ -1,806 +0,0 @@
-"""
-title: 导出为 Excel
-author: Fu-Jie
-author_url: https://github.com/Fu-Jie
-funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.3.3
-icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxwYXRoIGQ9Ik0xNSAySDZhMiAyIDAgMCAwLTIgMnYxNmEyIDIgMCAwIDAgMiAyaDEyYTIgMiAwIDAgMCAyLTJWN1oiLz48cGF0aCBkPSJNMTQgMnY0YTIgMiAwIDAgMCAyIDJoNCIvPjxwYXRoIGQ9Ik04IDEzaDIiLz48cGF0aCBkPSJNMTQgMTNoMiIvPjxwYXRoIGQ9Ik04IDE3aDIiLz48cGF0aCBkPSJNMTQgMTdoMiIvPjwvc3ZnPg==
-description: 将当前对话历史导出为 Excel (.xlsx) 文件，支持自动提取表头。
-"""
-
-import os
-import pandas as pd
-import re
-import base64
-from fastapi import FastAPI, HTTPException
-from typing import Optional, Callable, Awaitable, Any, List, Dict
-import datetime
-
-app = FastAPI()
-
-
-class Action:
-
-    def __init__(self):
-        pass
-
-    async def _send_notification(self, emitter: Callable, type: str, content: str):
-        await emitter(
-            {"type": "notification", "data": {"type": type, "content": content}}
-        )
-
-    async def action(
-        self,
-        body: dict,
-        __user__=None,
-        __event_emitter__=None,
-        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
-    ):
-        print(f"action:{__name__}")
-        if isinstance(__user__, (list, tuple)):
-            user_language = (
-                __user__[0].get("language", "zh-CN") if __user__ else "zh-CN"
-            )
-            user_name = __user__[0].get("name", "用户") if __user__[0] else "用户"
-            user_id = (
-                __user__[0]["id"]
-                if __user__ and "id" in __user__[0]
-                else "unknown_user"
-            )
-        elif isinstance(__user__, dict):
-            user_language = __user__.get("language", "zh-CN")
-            user_name = __user__.get("name", "用户")
-            user_id = __user__.get("id", "unknown_user")
-
-        if __event_emitter__:
-            last_assistant_message = body["messages"][-1]
-
-            await __event_emitter__(
-                {
-                    "type": "status",
-                    "data": {"description": "正在保存到文件...", "done": False},
-                }
-            )
-
-            try:
-                message_content = last_assistant_message["content"]
-                tables = self.extract_tables_from_message(message_content)
-
-                if not tables:
-                    raise HTTPException(status_code=400, detail="未找到任何表格。")
-
-                # 获取动态文件名和sheet名称
-                workbook_name, sheet_names = self.generate_names_from_content(
-                    message_content, tables
-                )
-
-                # 使用优化后的文件名生成逻辑
-                current_datetime = datetime.datetime.now()
-                formatted_date = current_datetime.strftime("%Y%m%d")
-
-                # 如果没找到标题则使用 user_yyyymmdd 格式
-                if not workbook_name:
-                    workbook_name = f"{user_name}_{formatted_date}"
-
-                filename = f"{workbook_name}.xlsx"
-                excel_file_path = os.path.join(
-                    "app", "backend", "data", "temp", filename
-                )
-
-                os.makedirs(os.path.dirname(excel_file_path), exist_ok=True)
-
-                # 保存表格到Excel（使用符合中国规范的格式化功能）
-                self.save_tables_to_excel_enhanced(tables, excel_file_path, sheet_names)
-
-                # 触发文件下载
-                if __event_call__:
-                    with open(excel_file_path, "rb") as file:
-                        file_content = file.read()
-                        base64_blob = base64.b64encode(file_content).decode("utf-8")
-
-                    await __event_call__(
-                        {
-                            "type": "execute",
-                            "data": {
-                                "code": f"""
-                                try {{
-                                    const base64Data = "{base64_blob}";
-                                    const binaryData = atob(base64Data);
-                                    const arrayBuffer = new Uint8Array(binaryData.length);
-                                    for (let i = 0; i < binaryData.length; i++) {{
-                                        arrayBuffer[i] = binaryData.charCodeAt(i);
-                                    }}
-                                    const blob = new Blob([arrayBuffer], {{ type: "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet" }});
-                                    const filename = "{filename}";
-
-                                    const url = URL.createObjectURL(blob);
-                                    const a = document.createElement("a");
-                                    a.style.display = "none";
-                                    a.href = url;
-                                    a.download = filename;
-                                    document.body.appendChild(a);
-                                    a.click();
-                                    URL.revokeObjectURL(url);
-                                    document.body.removeChild(a);
-                                }} catch (error) {{
-                                    console.error('触发下载时出错:', error);
-                                }}
-                                """
-                            },
-                        }
-                    )
-                await __event_emitter__(
-                    {
-                        "type": "status",
-                        "data": {"description": "输出已保存", "done": True},
-                    }
-                )
-
-                # 清理临时文件
-                if os.path.exists(excel_file_path):
-                    os.remove(excel_file_path)
-
-                return {"message": "下载事件已触发"}
-
-            except HTTPException as e:
-                print(f"Error processing tables: {str(e.detail)}")
-                await __event_emitter__(
-                    {
-                        "type": "status",
-                        "data": {
-                            "description": f"保存文件时出错: {e.detail}",
-                            "done": True,
-                        },
-                    }
-                )
-                await self._send_notification(
-                    __event_emitter__, "error", "没有找到可以导出的表格!"
-                )
-                raise e
-            except Exception as e:
-                print(f"Error processing tables: {str(e)}")
-                await __event_emitter__(
-                    {
-                        "type": "status",
-                        "data": {
-                            "description": f"保存文件时出错: {str(e)}",
-                            "done": True,
-                        },
-                    }
-                )
-                await self._send_notification(
-                    __event_emitter__, "error", "没有找到可以导出的表格!"
-                )
-
-    def extract_tables_from_message(self, message: str) -> List[Dict]:
-        """
-        从消息文本中提取Markdown表格及位置信息
-        返回结构: [{
-            "data": 表格数据,
-            "start_line": 起始行号,
-            "end_line": 结束行号
-        }]
-        """
-        table_row_pattern = r"^\s*\|.*\|.*\s*$"
-        rows = message.split("\n")
-        tables = []
-        current_table = []
-        start_line = None
-        current_line = 0
-
-        for row in rows:
-            current_line += 1
-            if re.search(table_row_pattern, row):
-                if start_line is None:
-                    start_line = current_line  # 记录表格起始行
-
-                # 处理表格行
-                cells = [cell.strip() for cell in row.strip().strip("|").split("|")]
-
-                # 跳过分隔行
-                is_separator_row = all(re.fullmatch(r"[:\-]+", cell) for cell in cells)
-                if not is_separator_row:
-                    current_table.append(cells)
-            elif current_table:
-                # 表格结束
-                tables.append(
-                    {
-                        "data": current_table,
-                        "start_line": start_line,
-                        "end_line": current_line - 1,
-                    }
-                )
-                current_table = []
-                start_line = None
-
-        # 处理最后一个表格
-        if current_table:
-            tables.append(
-                {
-                    "data": current_table,
-                    "start_line": start_line,
-                    "end_line": current_line,
-                }
-            )
-
-        return tables
-
-    def generate_names_from_content(self, content: str, tables: List[Dict]) -> tuple:
-        """
-        根据内容生成工作簿名称和sheet名称
-        - 忽略非空段落，只使用 markdown 标题 (h1-h6)。
-        - 单表格: 使用最近的标题作为工作簿和工作表名。
-        - 多表格: 使用文档第一个标题作为工作簿名，各表格最近的标题作为工作表名。
-        - 默认命名:
-            - 工作簿: 在主流程中处理 (user_yyyymmdd.xlsx)。
-            - 工作表: 表1, 表2, ...
-        """
-        lines = content.split("\n")
-        workbook_name = ""
-        sheet_names = []
-        all_headers = []
-
-        # 1. 查找文档中所有 h1-h6 标题及其位置
-        for i, line in enumerate(lines):
-            if re.match(r"^#{1,6}\s+", line):
-                all_headers.append(
-                    {"text": re.sub(r"^#{1,6}\s+", "", line).strip(), "line_num": i}
-                )
-
-        # 2. 为每个表格生成 sheet 名称
-        for i, table in enumerate(tables):
-            table_start_line = table["start_line"] - 1  # 转换为 0-based 索引
-            closest_header_text = None
-
-            # 查找当前表格上方最近的标题
-            candidate_headers = [
-                h for h in all_headers if h["line_num"] < table_start_line
-            ]
-            if candidate_headers:
-                # 找到候选标题中行号最大的，即为最接近的
-                closest_header = max(candidate_headers, key=lambda x: x["line_num"])
-                closest_header_text = closest_header["text"]
-
-            if closest_header_text:
-                # 清理并添加找到的标题
-                sheet_names.append(self.clean_sheet_name(closest_header_text))
-            else:
-                # 如果找不到标题，使用默认名称 "表{i+1}"
-                sheet_names.append(f"表{i+1}")
-
-        # 3. 根据表格数量确定工作簿名称
-        if len(tables) == 1:
-            # 单个表格: 使用其工作表名作为工作簿名 (前提是该名称不是默认的 "表1")
-            if sheet_names[0] != "表1":
-                workbook_name = sheet_names[0]
-        elif len(tables) > 1:
-            # 多个表格: 使用文档中的第一个标题作为工作簿名
-            if all_headers:
-                # 找到所有标题中行号最小的，即为第一个标题
-                first_header = min(all_headers, key=lambda x: x["line_num"])
-                workbook_name = first_header["text"]
-
-        # 4. 清理工作簿名称 (如果为空，主流程会使用默认名称)
-        workbook_name = self.clean_filename(workbook_name) if workbook_name else ""
-
-        return workbook_name, sheet_names
-
-    def clean_filename(self, name: str) -> str:
-        """清理文件名中的非法字符"""
-        return re.sub(r'[\\/*?:"<>|]', "", name).strip()
-
-    def clean_sheet_name(self, name: str) -> str:
-        """清理sheet名称(限制31字符,去除非法字符)"""
-        name = re.sub(r"[\\/*?[\]:]", "", name).strip()
-        return name[:31] if len(name) > 31 else name
-
-    # ======================== 符合中国规范的格式化功能 ========================
-
-    def calculate_text_width(self, text: str) -> float:
-        """
-        计算文本显示宽度，考虑中英文字符差异
-        中文字符按2个单位计算，英文字符按1个单位计算
-        """
-        if not text:
-            return 0
-
-        width = 0
-        for char in str(text):
-            # 判断是否为中文字符（包括中文标点）
-            if "\u4e00" <= char <= "\u9fff" or "\u3000" <= char <= "\u303f":
-                width += 2  # 中文字符占2个单位宽度
-            else:
-                width += 1  # 英文字符占1个单位宽度
-
-        return width
-
-    def calculate_text_height(self, text: str, max_width: int = 50) -> int:
-        """
-        计算文本显示所需的行数
-        根据换行符和文本长度计算
-        """
-        if not text:
-            return 1
-
-        text = str(text)
-        # 计算换行符导致的行数
-        explicit_lines = text.count("\n") + 1
-
-        # 计算因文本长度超出而需要的额外行数
-        text_width = self.calculate_text_width(text.replace("\n", ""))
-        wrapped_lines = max(
-            1, int(text_width / max_width) + (1 if text_width % max_width > 0 else 0)
-        )
-
-        return max(explicit_lines, wrapped_lines)
-
-    def get_column_letter(self, col_index: int) -> str:
-        """
-        将列索引转换为Excel列字母 (A, B, C, ..., AA, AB, ...)
-        """
-        result = ""
-        while col_index >= 0:
-            result = chr(65 + col_index % 26) + result
-            col_index = col_index // 26 - 1
-        return result
-
-    def determine_content_type(self, header: str, values: list) -> str:
-        """
-        根据表头和内容智能判断数据类型，符合中国官方表格规范
-        返回: 'number', 'date', 'sequence', 'text'
-        """
-        header_lower = str(header).lower().strip()
-
-        # 检查表头关键词
-        number_keywords = [
-            "数量",
-            "金额",
-            "价格",
-            "费用",
-            "成本",
-            "收入",
-            "支出",
-            "总计",
-            "小计",
-            "百分比",
-            "%",
-            "比例",
-            "率",
-            "数值",
-            "分数",
-            "成绩",
-            "得分",
-        ]
-        date_keywords = ["日期", "时间", "年份", "月份", "时刻", "date", "time"]
-        sequence_keywords = [
-            "序号",
-            "编号",
-            "号码",
-            "排序",
-            "次序",
-            "顺序",
-            "id",
-            "编码",
-        ]
-
-        # 检查表头
-        for keyword in number_keywords:
-            if keyword in header_lower:
-                return "number"
-
-        for keyword in date_keywords:
-            if keyword in header_lower:
-                return "date"
-
-        for keyword in sequence_keywords:
-            if keyword in header_lower:
-                return "sequence"
-
-        # 检查数据内容
-        if not values:
-            return "text"
-
-        sample_values = [
-            str(v).strip() for v in values[:10] if str(v).strip()
-        ]  # 取前10个非空值作为样本
-        if not sample_values:
-            return "text"
-
-        numeric_count = 0
-        date_count = 0
-        sequence_count = 0
-
-        for value in sample_values:
-            # 检查是否为数字
-            try:
-                float(
-                    value.replace(",", "")
-                    .replace("，", "")
-                    .replace("%", "")
-                    .replace("％", "")
-                )
-                numeric_count += 1
-                continue
-            except ValueError:
-                pass
-
-            # 检查是否为日期格式
-            date_patterns = [
-                r"\d{4}[-/年]\d{1,2}[-/月]\d{1,2}日?",
-                r"\d{1,2}[-/]\d{1,2}[-/]\d{4}",
-                r"\d{4}\d{2}\d{2}",
-            ]
-            for pattern in date_patterns:
-                if re.match(pattern, value):
-                    date_count += 1
-                    break
-
-            # 检查是否为序号格式
-            if (
-                re.match(r"^\d+$", value) and len(value) <= 4
-            ):  # 纯数字且不超过4位，可能是序号
-                sequence_count += 1
-
-        total_count = len(sample_values)
-
-        # 根据比例判断类型
-        if numeric_count / total_count >= 0.7:
-            return "number"
-        elif date_count / total_count >= 0.7:
-            return "date"
-        elif sequence_count / total_count >= 0.8 and sequence_count > 2:
-            return "sequence"
-        else:
-            return "text"
-
-    def get_column_letter(self, col_index: int) -> str:
-        """
-        将列索引转换为Excel列字母 (A, B, C, ..., AA, AB, ...)
-        """
-        result = ""
-        while col_index >= 0:
-            result = chr(65 + col_index % 26) + result
-            col_index = col_index // 26 - 1
-        return result
-
-    def save_tables_to_excel_enhanced(
-        self, tables: List[Dict], file_path: str, sheet_names: List[str]
-    ):
-        """
-        符合中国官方表格规范的Excel保存功能
-        """
-        try:
-            with pd.ExcelWriter(file_path, engine="xlsxwriter") as writer:
-                workbook = writer.book
-
-                # 定义表头样式 - 居中对齐（符合中国规范）
-                header_format = workbook.add_format(
-                    {
-                        "bold": True,
-                        "font_size": 12,
-                        "font_color": "white",
-                        "bg_color": "#00abbd",
-                        "border": 1,
-                        "align": "center",  # 表头居中
-                        "valign": "vcenter",
-                        "text_wrap": True,
-                    }
-                )
-
-                # 文本单元格样式 - 左对齐
-                text_format = workbook.add_format(
-                    {
-                        "border": 1,
-                        "align": "left",  # 文本左对齐
-                        "valign": "vcenter",
-                        "text_wrap": True,
-                    }
-                )
-
-                # 数值单元格样式 - 右对齐
-                number_format = workbook.add_format(
-                    {"border": 1, "align": "right", "valign": "vcenter"}  # 数值右对齐
-                )
-
-                # 整数格式 - 右对齐
-                integer_format = workbook.add_format(
-                    {
-                        "num_format": "0",
-                        "border": 1,
-                        "align": "right",  # 整数右对齐
-                        "valign": "vcenter",
-                    }
-                )
-
-                # 小数格式 - 右对齐
-                decimal_format = workbook.add_format(
-                    {
-                        "num_format": "0.00",
-                        "border": 1,
-                        "align": "right",  # 小数右对齐
-                        "valign": "vcenter",
-                    }
-                )
-
-                # 日期格式 - 居中对齐
-                date_format = workbook.add_format(
-                    {
-                        "border": 1,
-                        "align": "center",  # 日期居中对齐
-                        "valign": "vcenter",
-                        "text_wrap": True,
-                    }
-                )
-
-                # 序号格式 - 居中对齐
-                sequence_format = workbook.add_format(
-                    {
-                        "border": 1,
-                        "align": "center",  # 序号居中对齐
-                        "valign": "vcenter",
-                    }
-                )
-
-                for i, table in enumerate(tables):
-                    try:
-                        table_data = table["data"]
-                        if not table_data or len(table_data) < 1:
-                            print(f"Skipping empty table at index {i}")
-                            continue
-
-                        print(f"Processing table {i+1} with {len(table_data)} rows")
-
-                        # 获取sheet名称
-                        sheet_name = (
-                            sheet_names[i] if i < len(sheet_names) else f"表{i+1}"
-                        )
-
-                        # 创建DataFrame
-                        headers = [
-                            str(cell).strip()
-                            for cell in table_data[0]
-                            if str(cell).strip()
-                        ]
-                        if not headers:
-                            print(f"Warning: No valid headers found for table {i+1}")
-                            headers = [f"列{j+1}" for j in range(len(table_data[0]))]
-
-                        data_rows = []
-                        if len(table_data) > 1:
-                            max_cols = len(headers)
-                            for row in table_data[1:]:
-                                processed_row = []
-                                for j in range(max_cols):
-                                    if j < len(row):
-                                        processed_row.append(str(row[j]))
-                                    else:
-                                        processed_row.append("")
-                                data_rows.append(processed_row)
-                            df = pd.DataFrame(data_rows, columns=headers)
-                        else:
-                            df = pd.DataFrame(columns=headers)
-
-                        print(f"DataFrame created with columns: {list(df.columns)}")
-
-                        # 修复pandas FutureWarning - 使用try-except替代errors='ignore'
-                        for col in df.columns:
-                            try:
-                                df[col] = pd.to_numeric(df[col])
-                            except (ValueError, TypeError):
-                                pass
-
-                        # 先写入数据（不包含表头）
-                        df.to_excel(
-                            writer,
-                            sheet_name=sheet_name,
-                            index=False,
-                            header=False,
-                            startrow=1,
-                        )
-                        worksheet = writer.sheets[sheet_name]
-
-                        # 应用符合中国规范的格式化
-                        self.apply_chinese_standard_formatting(
-                            worksheet,
-                            df,
-                            headers,
-                            workbook,
-                            header_format,
-                            text_format,
-                            number_format,
-                            integer_format,
-                            decimal_format,
-                            date_format,
-                            sequence_format,
-                        )
-
-                    except Exception as e:
-                        print(f"Error processing table {i+1}: {str(e)}")
-                        continue
-
-        except Exception as e:
-            print(f"Error saving Excel file: {str(e)}")
-            raise
-
-    def apply_chinese_standard_formatting(
-        self,
-        worksheet,
-        df,
-        headers,
-        workbook,
-        header_format,
-        text_format,
-        number_format,
-        integer_format,
-        decimal_format,
-        date_format,
-        sequence_format,
-    ):
-        """
-        应用符合中国官方表格规范的格式化
-        - 表头: 居中对齐
-        - 数值: 右对齐
-        - 文本: 左对齐
-        - 日期: 居中对齐
-        - 序号: 居中对齐
-        """
-        try:
-            # 1. 写入表头（居中对齐）
-            print(f"Writing headers with Chinese standard alignment: {headers}")
-            for col_idx, header in enumerate(headers):
-                if header and str(header).strip():
-                    worksheet.write(0, col_idx, str(header).strip(), header_format)
-                else:
-                    default_header = f"列{col_idx+1}"
-                    worksheet.write(0, col_idx, default_header, header_format)
-
-            # 2. 分析每列的数据类型并应用相应格式
-            column_types = {}
-            for col_idx, column in enumerate(headers):
-                if col_idx < len(df.columns):
-                    column_values = df.iloc[:, col_idx].tolist()
-                    column_types[col_idx] = self.determine_content_type(
-                        column, column_values
-                    )
-                    print(
-                        f"Column '{column}' determined as type: {column_types[col_idx]}"
-                    )
-                else:
-                    column_types[col_idx] = "text"
-
-            # 3. 写入并格式化数据（根据类型使用不同对齐方式）
-            for row_idx, row in df.iterrows():
-                for col_idx, value in enumerate(row):
-                    content_type = column_types.get(col_idx, "text")
-
-                    # 根据内容类型选择格式
-                    if content_type == "number":
-                        # 数值类型 - 右对齐
-                        if pd.api.types.is_numeric_dtype(df.iloc[:, col_idx]):
-                            if pd.api.types.is_integer_dtype(df.iloc[:, col_idx]):
-                                current_format = integer_format
-                            else:
-                                try:
-                                    numeric_value = float(value)
-                                    if numeric_value.is_integer():
-                                        current_format = integer_format
-                                        value = int(numeric_value)
-                                    else:
-                                        current_format = decimal_format
-                                except (ValueError, TypeError):
-                                    current_format = decimal_format
-                        else:
-                            current_format = number_format
-
-                    elif content_type == "date":
-                        # 日期类型 - 居中对齐
-                        current_format = date_format
-
-                    elif content_type == "sequence":
-                        # 序号类型 - 居中对齐
-                        current_format = sequence_format
-
-                    else:
-                        # 文本类型 - 左对齐
-                        current_format = text_format
-
-                    worksheet.write(row_idx + 1, col_idx, value, current_format)
-
-            # 4. 自动调整列宽
-            for col_idx, column in enumerate(headers):
-                col_letter = self.get_column_letter(col_idx)
-
-                # 计算表头宽度
-                header_width = self.calculate_text_width(str(column))
-
-                # 计算数据列的最大宽度
-                max_data_width = 0
-                if not df.empty and col_idx < len(df.columns):
-                    for value in df.iloc[:, col_idx]:
-                        value_width = self.calculate_text_width(str(value))
-                        max_data_width = max(max_data_width, value_width)
-
-                # 基础宽度：取表头和数据的最大宽度
-                base_width = max(header_width, max_data_width)
-
-                # 根据内容类型调整宽度
-                content_type = column_types.get(col_idx, "text")
-                if content_type == "sequence":
-                    # 序号列通常比较窄
-                    optimal_width = max(8, min(15, base_width + 2))
-                elif content_type == "number":
-                    # 数值列需要额外空间显示数字
-                    optimal_width = max(12, min(25, base_width + 3))
-                elif content_type == "date":
-                    # 日期列需要固定宽度
-                    optimal_width = max(15, min(20, base_width + 2))
-                else:
-                    # 文本列根据内容调整
-                    if base_width <= 10:
-                        optimal_width = base_width + 3
-                    elif base_width <= 20:
-                        optimal_width = base_width + 4
-                    else:
-                        optimal_width = base_width + 5
-                    optimal_width = max(10, min(60, optimal_width))
-
-                worksheet.set_column(f"{col_letter}:{col_letter}", optimal_width)
-
-            # 5. 自动调整行高
-            # 设置表头行高为35点
-            worksheet.set_row(0, 35)
-
-            # 设置数据行行高
-            for row_idx, row in df.iterrows():
-                max_row_height = 20  # 中国表格规范建议的最小行高
-
-                for col_idx, value in enumerate(row):
-                    if col_idx < len(headers):
-                        col_width = min(
-                            60,
-                            max(
-                                10, self.calculate_text_width(str(headers[col_idx])) + 5
-                            ),
-                        )
-                    else:
-                        col_width = 15
-
-                    cell_lines = self.calculate_text_height(str(value), col_width)
-                    cell_height = cell_lines * 20  # 每行20点高度，符合中国规范
-
-                    max_row_height = max(max_row_height, cell_height)
-
-                final_height = min(120, max_row_height)
-                worksheet.set_row(row_idx + 1, final_height)
-
-            print(f"Successfully applied Chinese standard formatting")
-
-        except Exception as e:
-            print(f"Warning: Failed to apply Chinese standard formatting: {str(e)}")
-            # 降级到基础格式化
-            self.apply_basic_formatting_fallback(worksheet, df)
-
-    def apply_basic_formatting_fallback(self, worksheet, df):
-        """
-        基础格式化降级方案
-        """
-        try:
-            # 基础列宽调整
-            for i, column in enumerate(df.columns):
-                column_width = (
-                    max(
-                        len(str(column)),
-                        (df[column].astype(str).map(len).max() if not df.empty else 0),
-                    )
-                    + 2
-                )
-
-                col_letter = self.get_column_letter(i)
-                worksheet.set_column(
-                    f"{col_letter}:{col_letter}", min(60, max(10, column_width))
-                )
-
-            print("Applied basic formatting fallback")
-
-        except Exception as e:
-            print(f"Warning: Even basic formatting failed: {str(e)}")

plugins/actions/flash-card/README.md

@@ -48,3 +48,9 @@ GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
 ## License
 
 MIT License
+
+## Changelog
+
+### v0.2.4
+
+- Removed debug messages from output

plugins/actions/flash-card/README_CN.md

@@ -48,3 +48,9 @@ GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
 ## 许可证
 
 MIT License
+
+## 更新日志
+
+### v0.2.4
+
+- 移除输出中的调试信息

plugins/actions/flash-card/flash_card.py

@@ -3,7 +3,8 @@ title: Flash Card
 author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.2.2
+version: 0.2.4
+openwebui_id: 65a2ea8f-2a13-4587-9d76-55eea0035cc8
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxwb2x5Z29uIHBvaW50cz0iMTIgMiAyIDcgMTIgMTIgMjIgNyAxMiAyIi8+PHBvbHlsaW5lIHBvaW50cz0iMiAxNyAxMiAyMiAyMiAxNyIvPjxwb2x5bGluZSBwb2ludHM9IjIgMTIgMTIgMTcgMjIgMTIiLz48L3N2Zz4=
 description: Quickly generates beautiful flashcards from text, extracting key points and categories.
 """
@@ -147,7 +148,7 @@ class Action:
                     if role == "user"
                     else "Assistant" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} Message {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             return body

plugins/actions/flash-card/flash_card_cn.py

@@ -3,7 +3,8 @@ title: 闪记卡 (Flash Card)
 author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.2.2
+version: 0.2.4
+openwebui_id: 4a31eac3-a3c4-4c30-9ca5-dab36b5fac65
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxwb2x5Z29uIHBvaW50cz0iMTIgMiAyIDcgMTIgMTIgMjIgNyAxMiAyIi8+PHBvbHlsaW5lIHBvaW50cz0iMiAxNyAxMiAyMiAyMiAxNyIvPjxwb2x5bGluZSBwb2ludHM9IjIgMTIgMTIgMTcgMjIgMTIiLz48L3N2Zz4=
 description: 快速将文本提炼为精美的学习记忆卡片，支持核心要点提取与分类。
 """
@@ -144,7 +145,7 @@ class Action:
                     if role == "user"
                     else "助手" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} 消息 {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             return body

plugins/actions/infographic/README.md

@@ -1,25 +1,31 @@
 # 📊 Smart Infographic (AntV)
 
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.4.9 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
+
 An Open WebUI plugin powered by the AntV Infographic engine. It transforms long text into professional, beautiful infographics with a single click.
 
+## 🔥 What's New in v1.4.9
+
+- 🎨 **70+ Official Templates**: Integrated comprehensive AntV infographic template library.
+- 🖼️ **Iconify & unDraw Support**: Richer visuals with official icons and illustrations.
+- 📏 **Visual Optimization**: Improved text wrapping, adaptive sizing, and layout refinement.
+- ✨ **PNG Upload**: Infographics now upload as PNG format for better Word export compatibility.
+- 🔧 **Canvas Conversion**: Uses browser canvas for high-quality SVG to PNG conversion (2x scale).
+
+### Previous: v1.4.0
+
+- ✨ **Default Mode Change**: Default output mode is now `image` (static image) for better compatibility.
+- 📱 **Responsive Sizing**: Images now auto-adapt to the chat container width.
+
 ## ✨ Key Features
 
 - 🚀 **AI-Powered Transformation**: Automatically analyzes text logic, extracts key points, and generates structured charts.
-- 🎨 **Professional Templates**: Includes various AntV official templates: Lists, Trees, Mindmaps, Comparison Tables, Flowcharts, and Statistical Charts.
-- 🔍 **Auto-Icon Matching**: Built-in logic to search and match the most relevant Material Design Icons based on content.
+- 🎨 **70+ Professional Templates**: Includes various AntV official templates: Lists, Trees, Roadmaps, Timelines, Comparison Tables, SWOT, Quadrants, and Statistical Charts.
+- 🔍 **Auto-Icon Matching**: Built-in logic to search and match the most relevant icons (Iconify) and illustrations (unDraw).
 - 📥 **Multi-Format Export**: Download your infographics as **SVG**, **PNG**, or a **Standalone HTML** file.
 - 🌈 **Highly Customizable**: Supports Dark/Light modes, auto-adapts theme colors, with bold titles and refined card layouts.
 - 📱 **Responsive Design**: Generated charts look great on both desktop and mobile devices.
 
-## 🛠️ Supported Template Types
-
-| Category | Template Name | Use Case |
-| :--- | :--- | :--- |
-| **Lists & Hierarchy** | `list-grid`, `tree-vertical`, `mindmap` | Features, Org Charts, Brainstorming |
-| **Sequence & Relation** | `sequence-roadmap`, `relation-circle` | Roadmaps, Circular Flows, Steps |
-| **Comparison & Analysis** | `compare-binary`, `compare-swot`, `quadrant-quarter` | Pros/Cons, SWOT, Quadrants |
-| **Charts & Data** | `chart-bar`, `chart-line`, `chart-pie` | Trends, Distributions, Metrics |
-
 ## 🚀 How to Use
 
 1. **Install**: Search for "Smart Infographic" in the Open WebUI Community and install.
@@ -38,6 +44,17 @@ You can adjust the following parameters in the plugin settings to optimize the g
 | **Min Text Length (MIN_TEXT_LENGTH)** | `100` | Minimum characters required to trigger analysis, preventing accidental triggers on short text. |
 | **Clear Previous (CLEAR_PREVIOUS_HTML)** | `False` | Whether to clear previous charts. If `False`, new charts will be appended below. |
 | **Message Count (MESSAGE_COUNT)** | `1` | Number of recent messages to use for analysis. Increase this for more context. |
+| **Output Mode (OUTPUT_MODE)** | `image` | `image` for static image embedding (default, better compatibility), `html` for interactive chart. |
+
+## 🛠️ Supported Template Types
+
+| Category | Template Name | Use Case |
+| :--- | :--- | :--- |
+| **Sequence** | `sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-snake-steps-compact-card` | Timelines, Roadmaps, Processes |
+| **Lists** | `list-grid-candy-card-lite`, `list-row-horizontal-icon-arrow`, `list-column-simple-vertical-arrow` | Features, Bullet Points, Lists |
+| **Comparison** | `compare-binary-horizontal-underline-text-vs`, `compare-swot`, `quadrant-quarter-simple-card` | Pros/Cons, SWOT, Quadrants |
+| **Hierarchy** | `hierarchy-tree-tech-style-capsule-item`, `hierarchy-structure` | Org Charts, Structures |
+| **Charts** | `chart-column-simple`, `chart-bar-plain-text`, `chart-line-plain-text`, `chart-wordcloud` | Trends, Distributions, Metrics |
 
 ## 📝 Syntax Example (For Advanced Users)
 
@@ -54,12 +71,3 @@ data
     - label Beautiful Design
       desc Uses AntV professional design standards
 ```
-
-## 👨‍💻 Author
-
-**jeff**
-- GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## 📄 License
-
-MIT License

plugins/actions/infographic/README_CN.md

@@ -1,25 +1,31 @@
 # 📊 智能信息图 (AntV Infographic)
 
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.4.9 | **Project:** [Awesome OpenWebUI](https://github.com/Fu-Jie/awesome-openwebui)
+
 基于 AntV Infographic 引擎的 Open WebUI 插件，能够将长文本内容一键转换为专业、美观的信息图表。
 
+## 🔥 v1.4.9 更新日志
+
+- 🎨 **70+ 官方模板**：全面集成 AntV 官方信息图模板库。
+- 🖼️ **图标与插图支持**：支持 Iconify 图标库与 unDraw 插图库，视觉效果更丰富。
+- 📏 **视觉优化**：改进文本换行逻辑，优化自适应尺寸，提升卡片布局精细度。
+- ✨ **PNG 上传**：信息图现在以 PNG 格式上传，与 Word 导出完美兼容。
+- 🔧 **Canvas 转换**：使用浏览器 Canvas 高质量转换 SVG 为 PNG（2倍缩放）。
+
+### 此前: v1.4.0
+
+- ✨ **默认模式变更**：默认输出模式调整为 `image`（静态图片）。
+- 📱 **响应式尺寸**：图片模式下自动适应聊天容器宽度。
+
 ## ✨ 核心特性
 
 - 🚀 **智能转换**：自动分析文本核心逻辑，提取关键点并生成结构化图表。
-- 🎨 **专业模板**：内置多种 AntV 官方模板，包括列表、树图、思维导图、对比图、流程图及统计图表等。
-- 🔍 **自动图标匹配**：内置图标搜索逻辑，根据内容自动匹配最相关的 Material Design Icons。
+- 🎨 **70+ 专业模板**：内置多种 AntV 官方模板，包括列表、树图、路线图、时间线、对比图、SWOT、象限图及统计图表等。
+- 🔍 **自动图标匹配**：内置图标搜索逻辑，支持 Iconify 图标和 unDraw 插图自动匹配。
 - 📥 **多格式导出**：支持一键下载为 **SVG**、**PNG** 或 **独立 HTML** 文件。
 - 🌈 **高度自定义**：支持深色/浅色模式，自动适配主题颜色，主标题加粗突出，卡片布局精美。
 - 📱 **响应式设计**：生成的图表在桌面端和移动端均有良好的展示效果。
 
-## 🛠️ 支持的模板类型
-
-| 分类 | 模板名称 | 适用场景 |
-| :--- | :--- | :--- |
-| **列表与层级** | `list-grid`, `tree-vertical`, `mindmap` | 功能亮点、组织架构、思维导图 |
-| **顺序与关系** | `sequence-roadmap`, `relation-circle` | 发展历程、循环关系、步骤说明 |
-| **对比与分析** | `compare-binary`, `compare-swot`, `quadrant-quarter` | 优劣势对比、SWOT 分析、象限图 |
-| **图表与数据** | `chart-bar`, `chart-line`, `chart-pie` | 数据趋势、比例分布、数值对比 |
-
 ## 🚀 使用方法
 
 1. **安装插件**：在 Open WebUI 插件市场搜索并安装。
@@ -38,6 +44,17 @@
 | **最小文本长度 (MIN_TEXT_LENGTH)** | `100` | 触发分析所需的最小字符数，防止对过短的对话误操作。 |
 | **清除旧结果 (CLEAR_PREVIOUS_HTML)** | `False` | 每次生成是否清除之前的图表。若为 `False`，新图表将追加在下方。 |
 | **上下文消息数 (MESSAGE_COUNT)** | `1` | 用于分析的最近消息条数。增加此值可让 AI 参考更多对话背景。 |
+| **输出模式 (OUTPUT_MODE)** | `image` | `image` 为静态图片嵌入（默认，兼容性好），`html` 为交互式图表。 |
+
+## 🛠️ 支持的模板类型
+
+| 分类 | 模板名称 | 适用场景 |
+| :--- | :--- | :--- |
+| **时序与流程** | `sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-snake-steps-compact-card` | 时间线、路线图、步骤说明 |
+| **列表与网格** | `list-grid-candy-card-lite`, `list-row-horizontal-icon-arrow`, `list-column-simple-vertical-arrow` | 功能亮点、要点列举、清单 |
+| **对比与分析** | `compare-binary-horizontal-underline-text-vs`, `compare-swot`, `quadrant-quarter-simple-card` | 优劣势对比、SWOT 分析、象限图 |
+| **层级与结构** | `hierarchy-tree-tech-style-capsule-item`, `hierarchy-structure` | 组织架构、层级关系 |
+| **图表与数据** | `chart-column-simple`, `chart-bar-plain-text`, `chart-line-plain-text`, `chart-wordcloud` | 数据趋势、比例分布、数值对比 |
 
 ## 📝 语法示例 (高级用户)
 
@@ -54,12 +71,3 @@ data
     - label 视觉精美
       desc 采用 AntV 专业设计规范
 ```
-
-## 👨‍💻 作者
-
-**jeff**
-- GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## 📄 许可证
-
-MIT License

plugins/actions/infographic/infographic.py

@@ -1,14 +1,15 @@
 """
 title: 📊 Smart Infographic (AntV)
-author: jeff
+author: Fu-Jie
 author_url: https://github.com/Fu-Jie/awesome-openwebui
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPgogIDxsaW5lIHgxPSIxMiIgeTE9IjIwIiB4Mj0iMTIiIHkyPSIxMCIgLz4KICA8bGluZSB4MT0iMTgiIHkxPSIyMCIgeDI9IjE4IiB5Mj0iNCIgLz4KICA8bGluZSB4MT0iNiIgeTE9IjIwIiB4Mj0iNiIgeTI9IjE2IiAvPgo8L3N2Zz4=
-version: 1.3.0
+version: 1.4.9
+openwebui_id: ad6f0c7f-c571-4dea-821d-8e71697274cf
 description: AI-powered infographic generator based on AntV Infographic. Supports professional templates, auto-icon matching, and SVG/PNG downloads.
 """
 
 from pydantic import BaseModel, Field
-from typing import Optional, Dict, Any
+from typing import Optional, Dict, Any, Callable, Awaitable
 import logging
 import time
 import re
@@ -46,24 +47,63 @@ Infographic syntax is a Mermaid-like declarative syntax for describing infograph
 
 ### Template Library & Selection Guide
 
-Choose the most appropriate template based on the content structure:
+Choose the most appropriate template based on content structure.
 
-#### 1. List & Hierarchy
-- **List**: `list-grid` (Grid Cards), `list-vertical` (Vertical List)
-- **Tree**: `tree-vertical` (Vertical Tree), `tree-horizontal` (Horizontal Tree)
-- **Mindmap**: `mindmap` (Mind Map)
+**Template Selection Guidelines (Official):**
+- Strict sequential order (processes/steps/trends) → `sequence-*` series
+  - Timeline → `sequence-timeline-simple`
+  - Roadmap → `sequence-roadmap-vertical-simple`
+  - Zigzag steps → `sequence-horizontal-zigzag-underline-text`
+  - Snake steps → `sequence-snake-steps-compact-card`
+- Listing viewpoints → `list-row-horizontal-icon-arrow` or `list-column-simple-vertical-arrow`
+- Comparative analysis (A vs B) → `compare-binary-horizontal-underline-text-vs`
+- SWOT analysis → `compare-swot`
+- Hierarchical structure (tree) → `hierarchy-tree-tech-style-capsule-item`
+- Data charts → `chart-*` series
+- Quadrant analysis → `quadrant-quarter-simple-card`
+- Grid lists (bullet points) → `list-grid-candy-card-lite`
+- Relationship display → `relation-circle-icon-badge`
 
-#### 2. Sequence & Relationship
-- **Process**: `sequence-roadmap` (Roadmap), `sequence-zigzag` (Zigzag Process), `sequence-horizontal` (Horizontal Process)
-- **Relationship**: `relation-sankey` (Sankey Diagram), `relation-circle` (Circular Relationship)
+**Available Templates:**
 
-#### 3. Comparison & Analysis
-- **Comparison**: `compare-binary` (Binary Comparison), `list-grid` (Multi-item Grid Comparison)
-- **Analysis**: `compare-swot` (SWOT Analysis), `quadrant-quarter` (Quadrant Chart)
+*Sequence (时序/流程):*
+`sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-horizontal-zigzag-underline-text`, 
+`sequence-snake-steps-compact-card`, `sequence-zigzag-steps-underline-text`, `sequence-circular-simple`,
+`sequence-pyramid-simple`, `sequence-ascending-steps`
 
-#### 4. Charts & Data
-- **Statistics**: `statistic-card` (Statistic Cards)
-- **Charts**: `chart-bar` (Bar Chart), `chart-column` (Column Chart), `chart-line` (Line Chart), `chart-pie` (Pie Chart), `chart-doughnut` (Doughnut Chart), `chart-area` (Area Chart)
+*List (列表):*
+`list-grid-candy-card-lite`, `list-grid-badge-card`, `list-row-horizontal-icon-arrow`,
+`list-column-simple-vertical-arrow`, `list-column-done-list`
+
+*Compare (对比):*
+`compare-binary-horizontal-underline-text-vs`, `compare-binary-horizontal-simple-fold`,
+`compare-hierarchy-left-right-circle-node-pill-badge`, `compare-swot`
+
+*Hierarchy (层级):*
+`hierarchy-tree-tech-style-capsule-item`, `hierarchy-tree-curved-line-rounded-rect-node`, `hierarchy-structure`
+
+*Chart (图表):*
+`chart-column-simple`, `chart-bar-plain-text`, `chart-line-plain-text`, 
+`chart-pie-plain-text`, `chart-pie-donut-plain-text`, `chart-wordcloud`
+
+*Other:*
+`quadrant-quarter-simple-card`, `relation-circle-icon-badge`
+
+**Text Capacity by Template Type:**
+- HIGH capacity (long descriptions OK): `list-column-*`, `compare-binary-*`, `sequence-timeline-*`
+- MEDIUM capacity: `list-row-*`, `sequence-roadmap-*`
+- LOW capacity (short text only): `list-grid-*`, `hierarchy-*`, `sequence-steps`
+
+### Icon and Illustration Resources
+
+**Icons (Iconify):**
+- Format: `<collection>/<icon-name>`, e.g., `mdi/rocket-launch`
+- Popular: `mdi/*` (Material Design), `fa/*` (Font Awesome), `bi/*` (Bootstrap)
+- Examples: `mdi/code-tags`, `mdi/chart-line`, `mdi/account-group`, `mdi/cloud`
+
+**Illustrations (unDraw):**
+- Format: filename without .svg, e.g., `coding`, `team-work`
+- Use `illus` field instead of `icon`
 
 ### Data Structure Examples
 
@@ -210,6 +250,12 @@ data
 - `children`: Nested items (for trees, SWOT, etc.)
 - `illus`: Illustration icon (specific to some templates like Quadrant)
 
+### Content Refinement Principles
+1. **Brevity is King**: Infographics are visual. Keep text to a minimum.
+2. **Title Limit**: Keep `label` (item titles) under 15 characters (approx. 10 Chinese characters).
+3. **Description Limit**: Keep `desc` (item descriptions) under 40 characters (approx. 20 Chinese characters / 2 lines).
+4. **Impact**: Use strong verbs and nouns. Avoid filler words.
+
 ## Output Requirements
 1. **Language**: Output content in the user's language.
 2. **Format**: Wrap output in ```infographic ... ```.
@@ -232,9 +278,18 @@ User Language: {user_language}
 
 Please select the most appropriate infographic template based on text characteristics and output standard infographic syntax. Pay attention to correct indentation format (two spaces).
 
-**Important Note:** 
-- If using `list-grid` format, ensure each card's `desc` description is limited to **maximum 30 Chinese characters** (or **approximately 60 English characters**) to maintain visual consistency with all descriptions fitting in 2 lines.
-- Descriptions should be concise and highlight key points.
+**Visual Optimization Guide (MUST FOLLOW):**
+- **Point-based Generation:** Infographics are not articles. Extract KEYWORDS ONLY, avoid complete sentences.
+- **Main Title (`data.title`):** **MUST** be ≤ **15 Chinese characters** (or ≤30 English characters). Trim version numbers or details if needed.
+- **Subtitle (`data.desc`):** **MUST** be ≤ **20 Chinese characters** (or ≤40 English characters).
+- **Card Title (`label`):** **MUST** be ≤ **6 Chinese characters** (or ≤12 English characters). Use 2-4 keywords only.
+- **Card Description (`desc`):** **MUST** be ≤ **12 Chinese characters** (or ≤24 English characters). Use short phrases.
+
+⚠️ **CRITICAL**: If the original text is too long, you MUST rephrase and shorten it. Do NOT simply truncate with "...".
+Examples:
+- ❌ "多步任务与工具协作能力" → ✅ "多步任务协作"
+- ❌ "Open WebUI v0.7.x 重大版本更新" → ✅ "v0.7 核心更新"
+- ❌ "自动查找历史聊天记录" → ✅ "历史检索"
 """
 
 # =================================================================
@@ -339,8 +394,9 @@ CSS_TEMPLATE_INFOGRAPHIC = """
 .infographic-container-wrapper .infographic-render-container {
     border-radius: 8px;
     padding: 16px;
+    min-height: 600px;
     background: #fff;
-    overflow: visible; /* Ensure content is visible */
+    overflow: visible;
     transition: height 0.3s ease;
 }
 .infographic-render-container svg text {
@@ -348,35 +404,59 @@ CSS_TEMPLATE_INFOGRAPHIC = """
 }
 .infographic-render-container svg foreignObject {
     font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", sans-serif !important;
-    line-height: 1.4 !important;
+    line-height: 1.3 !important;
+    overflow: visible !important;
 }
 /* Main title styles */
 .infographic-render-container svg foreignObject[data-element-type="title"] > * {
-    font-size: 1.5em !important;
-    font-weight: bold !important;
-    line-height: 1.4 !important;
-    white-space: nowrap !important;
+    font-size: 1.3em !important;
+    font-weight: 800 !important;
+    line-height: 1.3 !important;
+    white-space: normal !important;
+    word-break: break-word !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
     overflow: hidden !important;
     text-overflow: ellipsis !important;
+    text-align: center !important;
 }
-/* Page subtitle and card title styles */
-.infographic-render-container svg foreignObject[data-element-type="desc"] > *,
-.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
-    font-size: 0.6em !important;
-    line-height: 1.4 !important;
-    white-space: nowrap !important;
-    overflow: hidden !important;
-    text-overflow: ellipsis !important;
-}
-/* Card title with extra bottom spacing */
-.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
-    padding-bottom: 8px !important;
+/* Page subtitle styles */
+.infographic-render-container svg foreignObject[data-element-type="desc"] > * {
+    font-size: 0.85em !important;
+    line-height: 1.3 !important;
+    white-space: normal !important;
+    word-break: break-word !important;
+    overflow: visible !important;
+    text-align: center !important;
     display: block !important;
+    color: var(--ig-muted-text-color) !important;
 }
-/* Card description text keeps normal wrapping */
+/* Card title styles */
+.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
+    font-size: 0.9em !important;
+    font-weight: 600 !important;
+    line-height: 1.3 !important;
+    white-space: normal !important;
+    word-break: break-word !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
+    padding-bottom: 2px !important;
+}
+/* Card description text */
 .infographic-render-container svg foreignObject[data-element-type="item-desc"] > * {
+    font-size: 0.8em !important;
     line-height: 1.4 !important;
     white-space: normal !important;
+    word-break: break-word !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
 }
 .infographic-container-wrapper .download-area {
     text-align: center;
@@ -532,37 +612,41 @@ SCRIPT_TEMPLATE_INFOGRAPHIC = """
             }}
         }}
 
-        // 2. Template Mapping Configuration
+        // 2. Template Mapping Configuration (Official AntV Structure IDs)
         const TEMPLATE_MAPPING = {{
-            // List & Hierarchy
+            // List & Hierarchy - map short names to full template names
             'list-grid': 'list-grid-compact-card',
+            'list-column': 'list-column-simple-vertical-arrow',
+            'list-row': 'list-row-simple-horizontal-arrow',
+            'hierarchy-tree': 'hierarchy-tree-tech-style-capsule-item',
+            
+            // Sequence & Timeline
+            'sequence-roadmap-vertical': 'sequence-roadmap-vertical-simple',
+            'sequence-timeline': 'sequence-timeline-simple',
+            'sequence-steps': 'sequence-steps-simple',
+            'sequence-horizontal-zigzag': 'sequence-horizontal-zigzag-simple',
+            
+            // Comparison
+            'compare-binary-horizontal': 'compare-binary-horizontal-simple-vs',
+            'compare-hierarchy-row': 'compare-hierarchy-row-simple',
+            
+            // Charts
+            'chart-column': 'chart-column-simple',
+            'quadrant': 'quadrant-quarter-simple-card',
+            
+            // Legacy mappings for backward compatibility
             'list-vertical': 'list-column-simple-vertical-arrow',
             'tree-vertical': 'hierarchy-tree-tech-style-capsule-item',
-            'tree-horizontal': 'hierarchy-tree-lr-tech-style-capsule-item',
-            'mindmap': 'hierarchy-mindmap-branch-gradient-capsule-item',
-            
-            // Sequence & Relationship
             'sequence-roadmap': 'sequence-roadmap-vertical-simple',
             'sequence-zigzag': 'sequence-horizontal-zigzag-simple',
-            'sequence-horizontal': 'sequence-horizontal-zigzag-simple',
-            'relation-sankey': 'relation-sankey-simple',
-            'relation-circle': 'relation-circle-icon-badge',
-            
-            // Comparison & Analysis
             'compare-binary': 'compare-binary-horizontal-simple-vs',
-            'compare-swot': 'compare-swot',
-            'quadrant-quarter': 'quadrant-quarter-simple-card',
-            
-            // Charts & Data
-            'statistic-card': 'list-grid-compact-card',
             'chart-bar': 'chart-bar-plain-text',
-            'chart-column': 'chart-column-simple',
             'chart-line': 'chart-line-plain-text',
-            'chart-area': 'chart-area-simple',
             'chart-pie': 'chart-pie-plain-text',
             'chart-doughnut': 'chart-pie-donut-plain-text'
         }};
 
+
         // 3. Apply Mapping Strategy
         for (const [key, value] of Object.entries(TEMPLATE_MAPPING)) {{
             const regex = new RegExp(`infographic\\\\s+${{key}}(?=\\\\s|$)`, 'i');
@@ -628,10 +712,48 @@ SCRIPT_TEMPLATE_INFOGRAPHIC = """
             containerEl.dataset.infographicRendered = 'true';
             console.log('[Infographic] Rendering complete');
 
-            // Auto-adjust height
+            // Auto-adjust height and tag elements
             setTimeout(() => {
                 const svg = containerEl.querySelector('svg');
                 if (svg) {
+                    // 1. Tag elements for CSS styling
+                    const fos = Array.from(svg.querySelectorAll('foreignObject'));
+                    let titleFound = false;
+                    let descFound = false;
+                    
+                    fos.forEach((fo) => {
+                        const text = fo.textContent.trim();
+                        if (!text || fo.querySelector('i') || (fo.querySelector('svg') && fo.querySelectorAll('*').length < 5)) {
+                            fo.setAttribute('data-element-type', 'icon');
+                            return;
+                        }
+                        
+                        // Dynamically increase height and width to accommodate wrapped text
+                        const currentHeight = parseInt(fo.getAttribute('height') || '0');
+                        if (currentHeight > 0 && currentHeight < 200) {
+                            fo.setAttribute('height', Math.round(currentHeight * 1.8).toString());
+                        }
+                        const currentWidth = parseInt(fo.getAttribute('width') || '0');
+                        if (currentWidth > 0 && currentWidth < 300) {
+                            fo.setAttribute('width', Math.max(Math.round(currentWidth * 1.2), 180).toString());
+                        }
+
+                        if (!titleFound) {
+                            fo.setAttribute('data-element-type', 'title');
+                            titleFound = true;
+                        } else if (!descFound) {
+                            fo.setAttribute('data-element-type', 'desc');
+                            descFound = true;
+                        } else {
+                            if (fo.querySelector('strong') || fo.style.fontWeight === 'bold' || text.length < 15) {
+                                fo.setAttribute('data-element-type', 'item-label');
+                            } else {
+                                fo.setAttribute('data-element-type', 'item-desc');
+                            }
+                        }
+                    });
+
+                    // 2. Adjust height
                     const bbox = svg.getBoundingClientRect();
                     let contentHeight = bbox.height;
                     if (svg.viewBox && svg.viewBox.baseVal && svg.viewBox.baseVal.height) {
@@ -821,10 +943,54 @@ class Action:
             default=1,
             description="Number of recent messages to use for generation. Set to 1 for just the last message, or higher for more context.",
         )
+        OUTPUT_MODE: str = Field(
+            default="image",
+            description="Output mode: 'html' for interactive HTML, or 'image' to embed as Markdown image (default).",
+        )
 
     def __init__(self):
         self.valves = self.Valves()
 
+    def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract chat_id from body or metadata"""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                chat_id = body_metadata.get("chat_id")
+                if isinstance(chat_id, str) and chat_id.strip():
+                    return chat_id.strip()
+
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+        return ""
+
+    def _extract_message_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract message_id from body or metadata"""
+        if isinstance(body, dict):
+            message_id = body.get("id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                message_id = body_metadata.get("message_id")
+                if isinstance(message_id, str) and message_id.strip():
+                    return message_id.strip()
+
+        if isinstance(metadata, dict):
+            message_id = metadata.get("message_id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+        return ""
+
     def _extract_infographic_syntax(self, llm_output: str) -> str:
         """Extract infographic syntax from LLM output"""
         match = re.search(r"```infographic\s*(.*?)\s*```", llm_output, re.DOTALL)
@@ -912,14 +1078,359 @@ class Action:
 
         return base_html.strip()
 
+    def _generate_image_js_code(
+        self,
+        unique_id: str,
+        chat_id: str,
+        message_id: str,
+        infographic_syntax: str,
+    ) -> str:
+        """Generate JavaScript code for frontend SVG rendering and image embedding"""
+
+        # Escape the syntax for JS embedding
+        syntax_escaped = (
+            infographic_syntax.replace("\\", "\\\\")
+            .replace("`", "\\`")
+            .replace("${", "\\${")
+            .replace("</script>", "<\\/script>")
+        )
+
+        return f"""
+(async function() {{
+    const uniqueId = "{unique_id}";
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    const defaultWidth = 1100;
+    const defaultHeight = 500;
+    
+    // Auto-detect chat container width for responsive sizing
+    let svgWidth = defaultWidth;
+    let svgHeight = defaultHeight;
+    const chatContainer = document.getElementById('chat-container');
+    if (chatContainer) {{
+        const containerWidth = chatContainer.clientWidth;
+        if (containerWidth > 100) {{
+            // Use container width with padding (80% of container, leaving more space on the right)
+            svgWidth = Math.floor(containerWidth * 0.8);
+            // Maintain aspect ratio based on default dimensions
+            svgHeight = Math.floor(svgWidth * (defaultHeight / defaultWidth));
+            console.log("[Infographic Image] Auto-detected container width:", containerWidth, "-> SVG:", svgWidth, "x", svgHeight);
+        }}
+    }}
+    
+    console.log("[Infographic Image] Starting render...");
+    console.log("[Infographic Image] chatId:", chatId, "messageId:", messageId);
+    
+    try {{
+        // Load AntV Infographic if not loaded
+        if (typeof AntVInfographic === 'undefined') {{
+            console.log("[Infographic Image] Loading AntV Infographic...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://unpkg.com/@antv/infographic@latest/dist/infographic.min.js';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        const {{ Infographic }} = AntVInfographic;
+        
+        // Get syntax content
+        let syntaxContent = `{syntax_escaped}`;
+        console.log("[Infographic Image] Syntax length:", syntaxContent.length);
+        
+        // Clean up syntax: remove code block markers
+        const backtick = String.fromCharCode(96);
+        const prefix = backtick + backtick + backtick + 'infographic';
+        const simplePrefix = backtick + backtick + backtick;
+        
+        if (syntaxContent.toLowerCase().startsWith(prefix)) {{
+            syntaxContent = syntaxContent.substring(prefix.length).trim();
+        }} else if (syntaxContent.startsWith(simplePrefix)) {{
+            syntaxContent = syntaxContent.substring(simplePrefix.length).trim();
+        }}
+        
+        if (syntaxContent.endsWith(simplePrefix)) {{
+            syntaxContent = syntaxContent.substring(0, syntaxContent.length - simplePrefix.length).trim();
+        }}
+        
+        // Fix syntax: remove colons after keywords
+        syntaxContent = syntaxContent.replace(/^(data|items|children|theme|config):/gm, '$1');
+        syntaxContent = syntaxContent.replace(/(\\s)(children|items):/g, '$1$2');
+        
+        // Ensure infographic prefix
+        if (!syntaxContent.trim().toLowerCase().startsWith('infographic')) {{
+            const firstWord = syntaxContent.trim().split(/\\s+/)[0].toLowerCase();
+            if (!['data', 'theme', 'design', 'items'].includes(firstWord)) {{
+                syntaxContent = 'infographic ' + syntaxContent;
+            }}
+        }}
+        
+        // Template mapping
+        const TEMPLATE_MAPPING = {{
+            'list-grid': 'list-grid-compact-card',
+            'list-vertical': 'list-column-simple-vertical-arrow',
+            'tree-vertical': 'hierarchy-tree-tech-style-capsule-item',
+            'tree-horizontal': 'hierarchy-tree-lr-tech-style-capsule-item',
+            'mindmap': 'hierarchy-mindmap-branch-gradient-capsule-item',
+            'sequence-roadmap': 'sequence-roadmap-vertical-simple',
+            'sequence-zigzag': 'sequence-horizontal-zigzag-simple',
+            'sequence-horizontal': 'sequence-horizontal-zigzag-simple',
+            'relation-sankey': 'relation-sankey-simple',
+            'relation-circle': 'relation-circle-icon-badge',
+            'compare-binary': 'compare-binary-horizontal-simple-vs',
+            'compare-swot': 'compare-swot',
+            'quadrant-quarter': 'quadrant-quarter-simple-card',
+            'statistic-card': 'list-grid-compact-card',
+            'chart-bar': 'chart-bar-plain-text',
+            'chart-column': 'chart-column-simple',
+            'chart-line': 'chart-line-plain-text',
+            'chart-area': 'chart-area-simple',
+            'chart-pie': 'chart-pie-plain-text',
+            'chart-doughnut': 'chart-pie-donut-plain-text'
+        }};
+        
+        for (const [key, value] of Object.entries(TEMPLATE_MAPPING)) {{
+            const regex = new RegExp(`infographic\\\\s+${{key}}(?=\\\\s|$)`, 'i');
+            if (regex.test(syntaxContent)) {{
+                syntaxContent = syntaxContent.replace(regex, `infographic ${{value}}`);
+                break;
+            }}
+        }}
+        
+        // Create offscreen container
+        const container = document.createElement('div');
+        container.id = 'infographic-offscreen-' + uniqueId;
+        container.style.cssText = 'position:absolute;left:-9999px;top:-9999px;width:' + svgWidth + 'px;height:' + svgHeight + 'px;background:#ffffff;';
+        document.body.appendChild(container);
+        
+        // Create infographic instance
+        const instance = new Infographic({{
+            container: '#' + container.id,
+            width: svgWidth,
+            height: svgHeight,
+            padding: 12,
+        }});
+        
+        console.log("[Infographic Image] Rendering infographic...");
+        instance.render(syntaxContent);
+        
+        // Wait for render to complete
+        await new Promise(resolve => setTimeout(resolve, 2000));
+        
+        // Get SVG element
+        const svgEl = container.querySelector('svg');
+        if (!svgEl) {{
+            throw new Error('SVG element not found after rendering');
+        }}
+        
+        // Get actual dimensions
+        const bbox = svgEl.getBoundingClientRect();
+        const width = bbox.width || svgWidth;
+        const height = bbox.height || svgHeight;
+        
+        // Clone and prepare SVG for export
+        const clonedSvg = svgEl.cloneNode(true);
+        clonedSvg.setAttribute('xmlns', 'http://www.w3.org/2000/svg');
+        clonedSvg.setAttribute('xmlns:xlink', 'http://www.w3.org/1999/xlink');
+        clonedSvg.setAttribute('width', width);
+        clonedSvg.setAttribute('height', height);
+        
+        // Add background rect
+        const bgRect = document.createElementNS('http://www.w3.org/2000/svg', 'rect');
+        bgRect.setAttribute('width', '100%');
+        bgRect.setAttribute('height', '100%');
+        bgRect.setAttribute('fill', '#ffffff');
+        clonedSvg.insertBefore(bgRect, clonedSvg.firstChild);
+        
+        // Serialize SVG to string
+        const svgData = new XMLSerializer().serializeToString(clonedSvg);
+        
+        // Cleanup container
+        document.body.removeChild(container);
+        
+        // Convert SVG to PNG using canvas for better compatibility
+        console.log("[Infographic Image] Converting SVG to PNG...");
+        const pngBlob = await new Promise((resolve, reject) => {{
+            const canvas = document.createElement('canvas');
+            const ctx = canvas.getContext('2d');
+            const scale = 2; // Higher resolution for clarity
+            canvas.width = Math.round(width * scale);
+            canvas.height = Math.round(height * scale);
+            
+            // Fill white background
+            ctx.fillStyle = '#ffffff';
+            ctx.fillRect(0, 0, canvas.width, canvas.height);
+            ctx.scale(scale, scale);
+            
+            const img = new Image();
+            img.onload = () => {{
+                ctx.drawImage(img, 0, 0, width, height);
+                canvas.toBlob((blob) => {{
+                    if (blob) {{
+                        resolve(blob);
+                    }} else {{
+                        reject(new Error('Canvas toBlob failed'));
+                    }}
+                }}, 'image/png');
+            }};
+            img.onerror = (e) => reject(new Error('Failed to load SVG as image: ' + e));
+            img.src = 'data:image/svg+xml;base64,' + btoa(unescape(encodeURIComponent(svgData)));
+        }});
+        
+        const file = new File([pngBlob], `infographic-${{uniqueId}}.png`, {{ type: 'image/png' }});
+        
+        // Upload file to OpenWebUI API
+        console.log("[Infographic Image] Uploading PNG file...");
+        const token = localStorage.getItem("token");
+        const formData = new FormData();
+        formData.append('file', file);
+        
+        const uploadResponse = await fetch('/api/v1/files/', {{
+            method: 'POST',
+            headers: {{
+                'Authorization': `Bearer ${{token}}`
+            }},
+            body: formData
+        }});
+        
+        if (!uploadResponse.ok) {{
+            throw new Error(`Upload failed: ${{uploadResponse.statusText}}`);
+        }}
+        
+        const fileData = await uploadResponse.json();
+        const fileId = fileData.id;
+        const imageUrl = `/api/v1/files/${{fileId}}/content`;
+        
+        console.log("[Infographic Image] PNG file uploaded, ID:", fileId);
+        
+        // Generate markdown image with file URL
+        const markdownImage = `![📊 Infographic](${{imageUrl}})`;
+        
+        // Update message via API
+        if (chatId && messageId) {{
+            
+            // Helper function with retry logic
+            const fetchWithRetry = async (url, options, retries = 3) => {{
+                for (let i = 0; i < retries; i++) {{
+                    try {{
+                        const response = await fetch(url, options);
+                        if (response.ok) return response;
+                        if (i < retries - 1) {{
+                            console.log(`[Infographic Image] Retry ${{i + 1}}/${{retries}} for ${{url}}`);
+                            await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                        }}
+                    }} catch (e) {{
+                        if (i === retries - 1) throw e;
+                        await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                    }}
+                }}
+                return null;
+            }};
+            
+            // Get current chat data
+            const getResponse = await fetch(`/api/v1/chats/${{chatId}}`, {{
+                method: "GET",
+                headers: {{ "Authorization": `Bearer ${{token}}` }}
+            }});
+            
+            if (!getResponse.ok) {{
+                throw new Error("Failed to get chat data: " + getResponse.status);
+            }}
+            
+            const chatData = await getResponse.json();
+            let updatedMessages = [];
+            let newContent = "";
+            
+            if (chatData.chat && chatData.chat.messages) {{
+                updatedMessages = chatData.chat.messages.map(m => {{
+                    if (m.id === messageId) {{
+                        const originalContent = m.content || "";
+                        // Remove existing infographic images
+                        const infographicPattern = /\\n*!\\[📊[^\\]]*\\]\\((?:data:image\\/[^)]+|(?:\\/api\\/v1\\/files\\/[^)]+))\\)/g;
+                        let cleanedContent = originalContent.replace(infographicPattern, "");
+                        cleanedContent = cleanedContent.replace(/\\n{{3,}}/g, "\\n\\n").trim();
+                        // Append new image
+                        newContent = cleanedContent + "\\n\\n" + markdownImage;
+                        
+                        // Update history object as well
+                        if (chatData.chat.history && chatData.chat.history.messages) {{
+                            if (chatData.chat.history.messages[messageId]) {{
+                                chatData.chat.history.messages[messageId].content = newContent;
+                            }}
+                        }}
+                        
+                        return {{ ...m, content: newContent }};
+                    }}
+                    return m;
+                }});
+            }}
+            
+            if (!newContent) {{
+                console.warn("[Infographic Image] Could not find message to update");
+                return;
+            }}
+            
+            // Try to update frontend display via event API
+            try {{
+                await fetch(`/api/v1/chats/${{chatId}}/messages/${{messageId}}/event`, {{
+                    method: "POST",
+                    headers: {{
+                        "Content-Type": "application/json",
+                        "Authorization": `Bearer ${{token}}`
+                    }},
+                    body: JSON.stringify({{
+                        type: "chat:message",
+                        data: {{ content: newContent }}
+                    }})
+                }});
+            }} catch (eventErr) {{
+                console.log("[Infographic Image] Event API not available, continuing...");
+            }}
+            
+            // Persist to database
+            const updatePayload = {{
+                chat: {{
+                    ...chatData.chat,
+                    messages: updatedMessages
+                }}
+            }};
+            
+            const persistResponse = await fetchWithRetry(`/api/v1/chats/${{chatId}}`, {{
+                method: "POST",
+                headers: {{
+                    "Content-Type": "application/json",
+                    "Authorization": `Bearer ${{token}}`
+                }},
+                body: JSON.stringify(updatePayload)
+            }});
+            
+            if (persistResponse && persistResponse.ok) {{
+                console.log("[Infographic Image] ✅ Message persisted successfully!");
+            }} else {{
+                console.error("[Infographic Image] ❌ Failed to persist message after retries");
+            }}
+        }} else {{
+            console.warn("[Infographic Image] ⚠️ Missing chatId or messageId, cannot persist");
+        }}
+        
+    }} catch (error) {{
+        console.error("[Infographic Image] Error:", error);
+    }}
+}})();
+"""
+
     async def action(
         self,
         body: dict,
         __user__: Optional[Dict[str, Any]] = None,
         __event_emitter__: Optional[Any] = None,
+        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __metadata__: Optional[dict] = None,
         __request__: Optional[Request] = None,
     ) -> Optional[dict]:
-        logger.info("Action: Infographic started (v1.0.0)")
+        logger.info("Action: Infographic started (v1.4.0)")
 
         # Get user information
         if isinstance(__user__, (list, tuple)):
@@ -961,9 +1472,7 @@ class Action:
                         if role == "user"
                         else "Assistant" if role == "assistant" else role
                     )
-                    aggregated_parts.append(
-                        f"[{role_label} Message {i}]\n{text_content}"
-                    )
+                    aggregated_parts.append(f"{text_content}")
 
             if not aggregated_parts:
                 raise ValueError("Unable to get valid user message content.")
@@ -1116,6 +1625,45 @@ class Action:
                         user_language,
                     )
 
+            # Check output mode
+            if self.valves.OUTPUT_MODE == "image":
+                # Image mode: use JavaScript to render and embed as Markdown image
+                chat_id = self._extract_chat_id(body, body.get("metadata"))
+                message_id = self._extract_message_id(body, body.get("metadata"))
+
+                await self._emit_status(
+                    __event_emitter__,
+                    "📊 Infographic: Rendering image...",
+                    False,
+                )
+
+                if __event_call__:
+                    js_code = self._generate_image_js_code(
+                        unique_id=unique_id,
+                        chat_id=chat_id,
+                        message_id=message_id,
+                        infographic_syntax=infographic_syntax,
+                    )
+
+                    await __event_call__(
+                        {
+                            "type": "execute",
+                            "data": {"code": js_code},
+                        }
+                    )
+
+                await self._emit_status(
+                    __event_emitter__, "✅ Infographic: Image generated!", True
+                )
+                await self._emit_notification(
+                    __event_emitter__,
+                    f"📊 Infographic image generated, {user_name}!",
+                    "success",
+                )
+                logger.info("Infographic generation completed in image mode")
+                return body
+
+            # HTML mode (default): embed as HTML block
             html_embed_tag = f"```html\n{final_html}\n```"
             body["messages"][-1]["content"] = f"{original_content}\n\n{html_embed_tag}"
 

plugins/actions/infographic/infographic_cn.py

@@ -1,14 +1,15 @@
 """
 title: 📊 智能信息图 (AntV Infographic)
-author: jeff
+author: Fu-Jie
 author_url: https://github.com/Fu-Jie/awesome-openwebui
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPgogIDxsaW5lIHgxPSIxMiIgeTE9IjIwIiB4Mj0iMTIiIHkyPSIxMCIgLz4KICA8bGluZSB4MT0iMTgiIHkxPSIyMCIgeDI9IjE4IiB5Mj0iNCIgLz4KICA8bGluZSB4MT0iNiIgeTE9IjIwIiB4Mj0iNiIgeTI9IjE2IiAvPgo8L3N2Zz4=
-version: 1.3.0
+version: 1.4.9
+openwebui_id: e04a48ff-23ee-4a41-8ea7-66c19524e7c8
 description: 基于 AntV Infographic 的智能信息图生成插件。支持多种专业模板，自动图标匹配，并提供 SVG/PNG 下载功能。
 """
 
 from pydantic import BaseModel, Field
-from typing import Optional, Dict, Any
+from typing import Optional, Dict, Any, Callable, Awaitable
 import logging
 import time
 import re
@@ -44,32 +45,61 @@ Infographic syntax is a Mermaid-like declarative syntax for describing infograph
 - ❌ Wrong: `children:` `items:` `data:` (with colons)
 - ✅ Correct: `children` `items` `data` (without colons)
 
-### Template Library & Selection Guide
+### 模板库与选择指南
 
-#### 1. List & Hierarchy (Text-heavy)
--   **Linear & Short (Steps/Phases)** -> `list-row-horizontal-icon-arrow`
--   **Linear & Long (Rankings/Details)** -> `list-vertical`
--   **Grouped / Parallel (Features/Catalog)** -> `list-grid`
--   **Hierarchical (Org Chart/Taxonomy)** -> `tree-vertical` or `tree-horizontal`
--   **Central Idea (Brainstorming)** -> `mindmap`
+根据内容结构选择最合适的模板。
 
-#### 2. Sequence & Relationship (Flow-based)
--   **Time-based (History/Plan)** -> `sequence-roadmap-vertical-simple`
--   **Process Flow (Complex)** -> `sequence-zigzag` or `sequence-horizontal`
--   **Resource Flow / Distribution** -> `relation-sankey`
--   **Circular Relationship** -> `relation-circle`
+**模板选择指南 (官方):**
+- 严格时序 (流程/步骤/趋势) → `sequence-*` 系列
+  - 时间线 → `sequence-timeline-simple`
+  - 路线图 → `sequence-roadmap-vertical-simple`
+  - 折线步骤 → `sequence-horizontal-zigzag-underline-text`
+  - 蛇形步骤 → `sequence-snake-steps-compact-card`
+- 列举要点 → `list-row-horizontal-icon-arrow` 或 `list-column-simple-vertical-arrow`
+- 对比分析 (A vs B) → `compare-binary-horizontal-underline-text-vs`
+- SWOT 分析 → `compare-swot`
+- 层级结构 (树状图) → `hierarchy-tree-tech-style-capsule-item`
+- 数据图表 → `chart-*` 系列
+- 象限分析 → `quadrant-quarter-simple-card`
+- 网格列表 → `list-grid-candy-card-lite`
+- 关系展示 → `relation-circle-icon-badge`
 
-#### 3. Comparison & Analysis
--   **Binary Comparison (A vs B)** -> `compare-binary`
--   **SWOT Analysis** -> `compare-swot`
--   **Quadrant Analysis (Importance vs Urgency)** -> `quadrant-quarter`
--   **Multi-item Grid Comparison** -> `list-grid` (use for comparing multiple items)
+**可用模板:**
 
-#### 4. Charts & Data (Metric-heavy)
--   **Key Metrics / Data Cards** -> `statistic-card`
--   **Distribution / Comparison** -> `chart-bar` or `chart-column`
--   **Trend over Time** -> `chart-line` or `chart-area`
--   **Proportion / Part-to-Whole** -> `chart-pie` or `chart-doughnut`
+*Sequence (时序/流程):*
+`sequence-timeline-simple`, `sequence-roadmap-vertical-simple`, `sequence-horizontal-zigzag-underline-text`, 
+`sequence-snake-steps-compact-card`, `sequence-zigzag-steps-underline-text`, `sequence-circular-simple`
+
+*List (列表):*
+`list-grid-candy-card-lite`, `list-grid-badge-card`, `list-row-horizontal-icon-arrow`,
+`list-column-simple-vertical-arrow`, `list-column-done-list`
+
+*Compare (对比):*
+`compare-binary-horizontal-underline-text-vs`, `compare-swot`
+
+*Hierarchy (层级):*
+`hierarchy-tree-tech-style-capsule-item`, `hierarchy-structure`
+
+*Chart (图表):*
+`chart-column-simple`, `chart-bar-plain-text`, `chart-pie-plain-text`, `chart-wordcloud`
+
+*Other:*
+`quadrant-quarter-simple-card`, `relation-circle-icon-badge`
+
+**按容量分类:**
+- 高容量 (长描述): `list-column-*`, `compare-binary-*`, `sequence-timeline-*`
+- 中容量: `list-row-*`, `sequence-roadmap-*`
+- 低容量 (短文本): `list-grid-*`, `hierarchy-*`
+
+### 图标和插图资源
+
+**图标 (Iconify):**
+- 格式: `<集合>/<图标名>`, 如 `mdi/rocket-launch`
+- 常用: `mdi/*`, `fa/*`, `bi/*`
+
+**插图 (unDraw):**
+- 格式: 文件名 (不含 .svg), 如 `coding`, `team-work`
+- 使用 `illus` 字段
 
 ### Infographic Syntax Guide
 
@@ -202,6 +232,12 @@ data
       desc Plan for next sprint
       illus mdi/star
 
+### Content Refinement Principles
+1. **Brevity is King**: Infographics are visual. Keep text to a minimum.
+2. **Title Limit**: Keep `label` (item titles) under 15 characters.
+3. **Description Limit**: Keep `desc` (item descriptions) under 25 characters (approx. 2 lines).
+4. **Impact**: Use strong verbs and nouns. Avoid filler words.
+
 ### Output Rules
 1.  **Strict Syntax**: Follow the indentation and formatting rules exactly.
 2.  **No Explanations**: Output ONLY the syntax code block.
@@ -223,9 +259,11 @@ USER_PROMPT_GENERATE_INFOGRAPHIC = """
 
 请根据文本特点选择最合适的信息图模板，并输出规范的 infographic 语法。注意保持正确的缩进格式（两个空格）。
 
-**重要提示：** 
-- 如果使用 `list-grid` 格式，请确保每个卡片的 `desc` 描述文字控制在 **30个汉字**（或约60个英文字符）**以内**，以保证所有卡片描述都只占用2行，维持视觉一致性。
-- 描述应简洁精炼，突出核心要点。
+**视觉优化指南：** 
+- **要点化生成：** 信息图不是文章。请将内容转化为“关键词+短语”的形式，严禁生成长难句。
+- **标题限制：** 每个卡片的 `label`（标题）请控制在 **8个汉字**以内。
+- **描述限制：** 每个卡片的 `desc`（描述）请控制在 **15个汉字**以内，确保即使在小屏幕上也能完整显示。
+- **结构化思维：** 优先使用并列、递进或对比结构，使信息一目了然。
 """
 
 # =================================================================
@@ -332,7 +370,7 @@ CSS_TEMPLATE_INFOGRAPHIC = """
     padding: 16px;
     min-height: 600px;
     background: #fff;
-    overflow: visible; /* Ensure content is visible */
+    overflow: visible;
     transition: height 0.3s ease;
 }
 .infographic-render-container svg text {
@@ -340,35 +378,58 @@ CSS_TEMPLATE_INFOGRAPHIC = """
 }
 .infographic-render-container svg foreignObject {
     font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", sans-serif !important;
-    line-height: 1.4 !important;
+    line-height: 1.3 !important;
+    overflow: visible !important;
 }
 /* 主标题样式 */
 .infographic-render-container svg foreignObject[data-element-type="title"] > * {
-    font-size: 1.5em !important;
-    font-weight: bold !important;
-    line-height: 1.4 !important;
-    white-space: nowrap !important;
-    overflow: hidden !important;
-    text-overflow: ellipsis !important;
-}
-/* 页面副标题和卡片标题样式 */
-.infographic-render-container svg foreignObject[data-element-type="desc"] > *,
-.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
-    font-size: 0.6em !important;
-    line-height: 1.4 !important;
-    white-space: nowrap !important;
-    overflow: hidden !important;
-    text-overflow: ellipsis !important;
-}
-/* 卡片标题额外增加底部间距 */
-.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
-    padding-bottom: 8px !important;
-    display: block !important;
-}
-/* 卡片描述文字保持正常换行 */
-.infographic-render-container svg foreignObject[data-element-type="item-desc"] > * {
-    line-height: 1.4 !important;
+    font-size: 1.3em !important;
+    font-weight: 800 !important;
+    line-height: 1.3 !important;
     white-space: normal !important;
+    word-break: break-word !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
+    text-align: center !important;
+}
+/* 页面副标题样式 */
+.infographic-render-container svg foreignObject[data-element-type="desc"] > * {
+    font-size: 0.85em !important;
+    line-height: 1.3 !important;
+    white-space: nowrap !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
+    text-align: center !important;
+    display: block !important;
+    color: var(--ig-muted-text-color) !important;
+}
+/* 卡片标题样式 */
+.infographic-render-container svg foreignObject[data-element-type="item-label"] > * {
+    font-size: 0.9em !important;
+    font-weight: 600 !important;
+    line-height: 1.3 !important;
+    white-space: normal !important;
+    word-break: break-word !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
+    padding-bottom: 2px !important;
+}
+/* 卡片描述文字 */
+.infographic-render-container svg foreignObject[data-element-type="item-desc"] > * {
+    font-size: 0.82em !important;
+    line-height: 1.3 !important;
+    white-space: normal !important;
+    display: -webkit-box !important;
+    -webkit-line-clamp: 2 !important;
+    -webkit-box-orient: vertical !important;
+    overflow: hidden !important;
+    text-overflow: ellipsis !important;
 }
 .infographic-container-wrapper .download-area {
     text-align: center;
@@ -536,34 +597,36 @@ SCRIPT_TEMPLATE_INFOGRAPHIC = """
             }
         }
 
-        // 2. 模板映射配置
-        // 2. 模板映射配置
+        // 2. 模板映射配置 (官方 AntV 结构 ID)
         const TEMPLATE_MAPPING = {
-            // 列表与层级
+            // 列表与层级 - 短名称映射到完整模板名
             'list-grid': 'list-grid-compact-card',
+            'list-column': 'list-column-simple-vertical-arrow',
+            'list-row': 'list-row-simple-horizontal-arrow',
+            'hierarchy-tree': 'hierarchy-tree-tech-style-capsule-item',
+            
+            // 时序与时间线
+            'sequence-roadmap-vertical': 'sequence-roadmap-vertical-simple',
+            'sequence-timeline': 'sequence-timeline-simple',
+            'sequence-steps': 'sequence-steps-simple',
+            'sequence-horizontal-zigzag': 'sequence-horizontal-zigzag-simple',
+            
+            // 对比
+            'compare-binary-horizontal': 'compare-binary-horizontal-simple-vs',
+            'compare-hierarchy-row': 'compare-hierarchy-row-simple',
+            
+            // 图表
+            'chart-column': 'chart-column-simple',
+            'quadrant': 'quadrant-quarter-simple-card',
+            
+            // 向后兼容的旧映射
             'list-vertical': 'list-column-simple-vertical-arrow',
             'tree-vertical': 'hierarchy-tree-tech-style-capsule-item',
-            'tree-horizontal': 'hierarchy-tree-lr-tech-style-capsule-item',
-            'mindmap': 'hierarchy-mindmap-branch-gradient-capsule-item',
-            
-            // 顺序与关系
             'sequence-roadmap': 'sequence-roadmap-vertical-simple',
             'sequence-zigzag': 'sequence-horizontal-zigzag-simple',
-            'sequence-horizontal': 'sequence-horizontal-zigzag-simple',
-            'relation-sankey': 'relation-sankey-simple', // 暂无直接对应，保留原值或需移除
-            'relation-circle': 'relation-circle-icon-badge',
-            
-            // 对比与分析
             'compare-binary': 'compare-binary-horizontal-simple-vs',
-            'compare-swot': 'compare-swot',
-            'quadrant-quarter': 'quadrant-quarter-simple-card',
-            
-            // 图表与数据
-            'statistic-card': 'list-grid-compact-card',
             'chart-bar': 'chart-bar-plain-text',
-            'chart-column': 'chart-column-simple',
             'chart-line': 'chart-line-plain-text',
-            'chart-area': 'chart-area-simple', // 暂无直接对应
             'chart-pie': 'chart-pie-plain-text',
             'chart-doughnut': 'chart-pie-donut-plain-text'
         };
@@ -656,10 +719,48 @@ SCRIPT_TEMPLATE_INFOGRAPHIC = """
             containerEl.dataset.infographicRendered = 'true';
             console.log('[Infographic] 渲染完成');
 
-            // 自动调整高度
+            // 自动调整高度与元素标记
             setTimeout(() => {
                 const svg = containerEl.querySelector('svg');
                 if (svg) {
+                    // 1. 标记元素以便 CSS 应用样式
+                    const fos = Array.from(svg.querySelectorAll('foreignObject'));
+                    let titleFound = false;
+                    let descFound = false;
+                    
+                    fos.forEach((fo) => {
+                        const text = fo.textContent.trim();
+                        if (!text || fo.querySelector('i') || (fo.querySelector('svg') && fo.querySelectorAll('*').length < 5)) {
+                            fo.setAttribute('data-element-type', 'icon');
+                            return;
+                        }
+                        
+                        // 动态增加高度和宽度，容纳换行后的文字
+                        const currentHeight = parseInt(fo.getAttribute('height') || '0');
+                        if (currentHeight > 0 && currentHeight < 200) {
+                            fo.setAttribute('height', Math.round(currentHeight * 1.8).toString());
+                        }
+                        const currentWidth = parseInt(fo.getAttribute('width') || '0');
+                        if (currentWidth > 0 && currentWidth < 300) {
+                            fo.setAttribute('width', Math.max(Math.round(currentWidth * 1.2), 180).toString());
+                        }
+
+                        if (!titleFound) {
+                            fo.setAttribute('data-element-type', 'title');
+                            titleFound = true;
+                        } else if (!descFound) {
+                            fo.setAttribute('data-element-type', 'desc');
+                            descFound = true;
+                        } else {
+                            if (fo.querySelector('strong') || fo.style.fontWeight === 'bold' || text.length < 15) {
+                                fo.setAttribute('data-element-type', 'item-label');
+                            } else {
+                                fo.setAttribute('data-element-type', 'item-desc');
+                            }
+                        }
+                    });
+
+                    // 2. 调整高度
                     const bbox = svg.getBoundingClientRect();
                     let contentHeight = bbox.height;
                     if (svg.viewBox && svg.viewBox.baseVal && svg.viewBox.baseVal.height) {
@@ -849,6 +950,10 @@ class Action:
             default=1,
             description="用于生成的最近消息数量。设置为1仅使用最后一条消息，更大值可包含更多上下文。",
         )
+        OUTPUT_MODE: str = Field(
+            default="image",
+            description="输出模式：'html' 为交互式HTML，'image' 将嵌入为Markdown图片（默认）。",
+        )
 
     def __init__(self):
         self.valves = self.Valves()
@@ -862,6 +967,46 @@ class Action:
             "Sunday": "星期日",
         }
 
+    def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """从 body 或 metadata 中提取 chat_id"""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                chat_id = body_metadata.get("chat_id")
+                if isinstance(chat_id, str) and chat_id.strip():
+                    return chat_id.strip()
+
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+        return ""
+
+    def _extract_message_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """从 body 或 metadata 中提取 message_id"""
+        if isinstance(body, dict):
+            message_id = body.get("id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                message_id = body_metadata.get("message_id")
+                if isinstance(message_id, str) and message_id.strip():
+                    return message_id.strip()
+
+        if isinstance(metadata, dict):
+            message_id = metadata.get("message_id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+        return ""
+
     def _extract_infographic_syntax(self, llm_output: str) -> str:
         """提取LLM输出中的infographic语法"""
         # 1. 优先匹配 ```infographic
@@ -973,14 +1118,359 @@ class Action:
 
         return base_html.strip()
 
+    def _generate_image_js_code(
+        self,
+        unique_id: str,
+        chat_id: str,
+        message_id: str,
+        infographic_syntax: str,
+    ) -> str:
+        """生成前端 SVG 渲染和图片嵌入的 JavaScript 代码"""
+
+        # 转义语法以便在 JS 中嵌入
+        syntax_escaped = (
+            infographic_syntax.replace("\\", "\\\\")
+            .replace("`", "\\`")
+            .replace("${", "\\${")
+            .replace("</script>", "<\\/script>")
+        )
+
+        return f"""
+(async function() {{
+    const uniqueId = "{unique_id}";
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    const defaultWidth = 1100;
+    const defaultHeight = 500;
+    
+    // 自动检测聊天容器宽度以实现响应式尺寸
+    let svgWidth = defaultWidth;
+    let svgHeight = defaultHeight;
+    const chatContainer = document.getElementById('chat-container');
+    if (chatContainer) {{
+        const containerWidth = chatContainer.clientWidth;
+        if (containerWidth > 100) {{
+            // 使用容器宽度的 80%（右边留更多空间）
+            svgWidth = Math.floor(containerWidth * 0.8);
+            // 根据默认尺寸保持宽高比
+            svgHeight = Math.floor(svgWidth * (defaultHeight / defaultWidth));
+            console.log("[Infographic Image] 自动检测容器宽度:", containerWidth, "-> SVG:", svgWidth, "x", svgHeight);
+        }}
+    }}
+    
+    console.log("[Infographic Image] 开始渲染...");
+    console.log("[Infographic Image] chatId:", chatId, "messageId:", messageId);
+    
+    try {{
+        // 加载 AntV Infographic（如果未加载）
+        if (typeof AntVInfographic === 'undefined') {{
+            console.log("[Infographic Image] 加载 AntV Infographic...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://registry.npmmirror.com/@antv/infographic/0.2.1/files/dist/infographic.min.js';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        const {{ Infographic }} = AntVInfographic;
+        
+        // 获取语法内容
+        let syntaxContent = `{syntax_escaped}`;
+        console.log("[Infographic Image] 语法长度:", syntaxContent.length);
+        
+        // 清理语法：移除代码块标记
+        const backtick = String.fromCharCode(96);
+        const prefix = backtick + backtick + backtick + 'infographic';
+        const simplePrefix = backtick + backtick + backtick;
+        
+        if (syntaxContent.toLowerCase().startsWith(prefix)) {{
+            syntaxContent = syntaxContent.substring(prefix.length).trim();
+        }} else if (syntaxContent.startsWith(simplePrefix)) {{
+            syntaxContent = syntaxContent.substring(simplePrefix.length).trim();
+        }}
+        
+        if (syntaxContent.endsWith(simplePrefix)) {{
+            syntaxContent = syntaxContent.substring(0, syntaxContent.length - simplePrefix.length).trim();
+        }}
+        
+        // 修复语法：移除关键字后的冒号
+        syntaxContent = syntaxContent.replace(/^(data|items|children|theme|config):/gm, '$1');
+        syntaxContent = syntaxContent.replace(/(\\s)(children|items):/g, '$1$2');
+        
+        // 确保 infographic 前缀
+        if (!syntaxContent.trim().toLowerCase().startsWith('infographic')) {{
+            const firstWord = syntaxContent.trim().split(/\\s+/)[0].toLowerCase();
+            if (!['data', 'theme', 'design', 'items'].includes(firstWord)) {{
+                syntaxContent = 'infographic ' + syntaxContent;
+            }}
+        }}
+        
+        // 模板映射
+        const TEMPLATE_MAPPING = {{
+            'list-grid': 'list-grid-compact-card',
+            'list-vertical': 'list-column-simple-vertical-arrow',
+            'tree-vertical': 'hierarchy-tree-tech-style-capsule-item',
+            'tree-horizontal': 'hierarchy-tree-lr-tech-style-capsule-item',
+            'mindmap': 'hierarchy-mindmap-branch-gradient-capsule-item',
+            'sequence-roadmap': 'sequence-roadmap-vertical-simple',
+            'sequence-zigzag': 'sequence-horizontal-zigzag-simple',
+            'sequence-horizontal': 'sequence-horizontal-zigzag-simple',
+            'relation-sankey': 'relation-sankey-simple',
+            'relation-circle': 'relation-circle-icon-badge',
+            'compare-binary': 'compare-binary-horizontal-simple-vs',
+            'compare-swot': 'compare-swot',
+            'quadrant-quarter': 'quadrant-quarter-simple-card',
+            'statistic-card': 'list-grid-compact-card',
+            'chart-bar': 'chart-bar-plain-text',
+            'chart-column': 'chart-column-simple',
+            'chart-line': 'chart-line-plain-text',
+            'chart-area': 'chart-area-simple',
+            'chart-pie': 'chart-pie-plain-text',
+            'chart-doughnut': 'chart-pie-donut-plain-text'
+        }};
+        
+        for (const [key, value] of Object.entries(TEMPLATE_MAPPING)) {{
+            const regex = new RegExp(`infographic\\\\s+${{key}}(?=\\\\s|$)`, 'i');
+            if (regex.test(syntaxContent)) {{
+                syntaxContent = syntaxContent.replace(regex, `infographic ${{value}}`);
+                break;
+            }}
+        }}
+        
+        // 创建离屏容器
+        const container = document.createElement('div');
+        container.id = 'infographic-offscreen-' + uniqueId;
+        container.style.cssText = 'position:absolute;left:-9999px;top:-9999px;width:' + svgWidth + 'px;height:' + svgHeight + 'px;background:#ffffff;';
+        document.body.appendChild(container);
+        
+        // 创建信息图实例
+        const instance = new Infographic({{
+            container: '#' + container.id,
+            width: svgWidth,
+            height: svgHeight,
+            padding: 12,
+        }});
+        
+        console.log("[Infographic Image] 渲染信息图...");
+        instance.render(syntaxContent);
+        
+        // 等待渲染完成
+        await new Promise(resolve => setTimeout(resolve, 2000));
+        
+        // 获取 SVG 元素
+        const svgEl = container.querySelector('svg');
+        if (!svgEl) {{
+            throw new Error('渲染后未找到 SVG 元素');
+        }}
+        
+        // 获取实际尺寸
+        const bbox = svgEl.getBoundingClientRect();
+        const width = bbox.width || svgWidth;
+        const height = bbox.height || svgHeight;
+        
+        // 克隆并准备导出的 SVG
+        const clonedSvg = svgEl.cloneNode(true);
+        clonedSvg.setAttribute('xmlns', 'http://www.w3.org/2000/svg');
+        clonedSvg.setAttribute('xmlns:xlink', 'http://www.w3.org/1999/xlink');
+        clonedSvg.setAttribute('width', width);
+        clonedSvg.setAttribute('height', height);
+        
+        // 添加背景矩形
+        const bgRect = document.createElementNS('http://www.w3.org/2000/svg', 'rect');
+        bgRect.setAttribute('width', '100%');
+        bgRect.setAttribute('height', '100%');
+        bgRect.setAttribute('fill', '#ffffff');
+        clonedSvg.insertBefore(bgRect, clonedSvg.firstChild);
+        
+        // 序列化 SVG 为字符串
+        const svgData = new XMLSerializer().serializeToString(clonedSvg);
+        
+        // 清理容器
+        document.body.removeChild(container);
+        
+        // 使用 canvas 将 SVG 转换为 PNG 以提高兼容性
+        console.log("[Infographic Image] 正在将 SVG 转换为 PNG...");
+        const pngBlob = await new Promise((resolve, reject) => {{
+            const canvas = document.createElement('canvas');
+            const ctx = canvas.getContext('2d');
+            const scale = 2; // 更高分辨率以提高清晰度
+            canvas.width = Math.round(width * scale);
+            canvas.height = Math.round(height * scale);
+            
+            // 填充白色背景
+            ctx.fillStyle = '#ffffff';
+            ctx.fillRect(0, 0, canvas.width, canvas.height);
+            ctx.scale(scale, scale);
+            
+            const img = new Image();
+            img.onload = () => {{
+                ctx.drawImage(img, 0, 0, width, height);
+                canvas.toBlob((blob) => {{
+                    if (blob) {{
+                        resolve(blob);
+                    }} else {{
+                        reject(new Error('Canvas toBlob 失败'));
+                    }}
+                }}, 'image/png');
+            }};
+            img.onerror = (e) => reject(new Error('加载 SVG 图片失败: ' + e));
+            img.src = 'data:image/svg+xml;base64,' + btoa(unescape(encodeURIComponent(svgData)));
+        }});
+        
+        const file = new File([pngBlob], `infographic-${{uniqueId}}.png`, {{ type: 'image/png' }});
+        
+        // 上传文件到 OpenWebUI API
+        console.log("[Infographic Image] 上传 PNG 文件...");
+        const token = localStorage.getItem("token");
+        const formData = new FormData();
+        formData.append('file', file);
+        
+        const uploadResponse = await fetch('/api/v1/files/', {{
+            method: 'POST',
+            headers: {{
+                'Authorization': `Bearer ${{token}}`
+            }},
+            body: formData
+        }});
+        
+        if (!uploadResponse.ok) {{
+            throw new Error(`上传失败: ${{uploadResponse.statusText}}`);
+        }}
+        
+        const fileData = await uploadResponse.json();
+        const fileId = fileData.id;
+        const imageUrl = `/api/v1/files/${{fileId}}/content`;
+        
+        console.log("[Infographic Image] PNG 文件已上传, ID:", fileId);
+        
+        // 生成带文件 URL 的 markdown 图片
+        const markdownImage = `![📊 信息图](${{imageUrl}})`;
+        
+        // 通过 API 更新消息
+        if (chatId && messageId) {{
+            
+            // 带重试逻辑的辅助函数
+            const fetchWithRetry = async (url, options, retries = 3) => {{
+                for (let i = 0; i < retries; i++) {{
+                    try {{
+                        const response = await fetch(url, options);
+                        if (response.ok) return response;
+                        if (i < retries - 1) {{
+                            console.log(`[Infographic Image] 重试 ${{i + 1}}/${{retries}} for ${{url}}`);
+                            await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                        }}
+                    }} catch (e) {{
+                        if (i === retries - 1) throw e;
+                        await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                    }}
+                }}
+                return null;
+            }};
+            
+            // 获取当前聊天数据
+            const getResponse = await fetch(`/api/v1/chats/${{chatId}}`, {{
+                method: "GET",
+                headers: {{ "Authorization": `Bearer ${{token}}` }}
+            }});
+            
+            if (!getResponse.ok) {{
+                throw new Error("获取聊天数据失败: " + getResponse.status);
+            }}
+            
+            const chatData = await getResponse.json();
+            let updatedMessages = [];
+            let newContent = "";
+            
+            if (chatData.chat && chatData.chat.messages) {{
+                updatedMessages = chatData.chat.messages.map(m => {{
+                    if (m.id === messageId) {{
+                        const originalContent = m.content || "";
+                        // 移除已有的信息图图片
+                        const infographicPattern = /\\n*!\\[📊[^\\]]*\\]\\((?:data:image\\/[^)]+|(?:\\/api\\/v1\\/files\\/[^)]+))\\)/g;
+                        let cleanedContent = originalContent.replace(infographicPattern, "");
+                        cleanedContent = cleanedContent.replace(/\\n{{3,}}/g, "\\n\\n").trim();
+                        // 追加新图片
+                        newContent = cleanedContent + "\\n\\n" + markdownImage;
+                        
+                        // 同时更新 history 对象
+                        if (chatData.chat.history && chatData.chat.history.messages) {{
+                            if (chatData.chat.history.messages[messageId]) {{
+                                chatData.chat.history.messages[messageId].content = newContent;
+                            }}
+                        }}
+                        
+                        return {{ ...m, content: newContent }};
+                    }}
+                    return m;
+                }});
+            }}
+            
+            if (!newContent) {{
+                console.warn("[Infographic Image] 找不到要更新的消息");
+                return;
+            }}
+            
+            // 尝试通过事件 API 更新前端显示
+            try {{
+                await fetch(`/api/v1/chats/${{chatId}}/messages/${{messageId}}/event`, {{
+                    method: "POST",
+                    headers: {{
+                        "Content-Type": "application/json",
+                        "Authorization": `Bearer ${{token}}`
+                    }},
+                    body: JSON.stringify({{
+                        type: "chat:message",
+                        data: {{ content: newContent }}
+                    }})
+                }});
+            }} catch (eventErr) {{
+                console.log("[Infographic Image] 事件 API 不可用，继续...");
+            }}
+            
+            // 持久化到数据库
+            const updatePayload = {{
+                chat: {{
+                    ...chatData.chat,
+                    messages: updatedMessages
+                }}
+            }};
+            
+            const persistResponse = await fetchWithRetry(`/api/v1/chats/${{chatId}}`, {{
+                method: "POST",
+                headers: {{
+                    "Content-Type": "application/json",
+                    "Authorization": `Bearer ${{token}}`
+                }},
+                body: JSON.stringify(updatePayload)
+            }});
+            
+            if (persistResponse && persistResponse.ok) {{
+                console.log("[Infographic Image] ✅ 消息持久化成功!");
+            }} else {{
+                console.error("[Infographic Image] ❌ 重试后消息持久化失败");
+            }}
+        }} else {{
+            console.warn("[Infographic Image] ⚠️ 缺少 chatId 或 messageId，无法持久化");
+        }}
+        
+    }} catch (error) {{
+        console.error("[Infographic Image] 错误:", error);
+    }}
+}})();
+"""
+
     async def action(
         self,
         body: dict,
         __user__: Optional[Dict[str, Any]] = None,
         __event_emitter__: Optional[Any] = None,
+        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __metadata__: Optional[dict] = None,
         __request__: Optional[Request] = None,
     ) -> Optional[dict]:
-        logger.info("Action: 信息图启动 (v1.0.0)")
+        logger.info("Action: 信息图启动 (v1.4.0)")
 
         # 获取用户信息
         if isinstance(__user__, (list, tuple)):
@@ -1026,7 +1516,7 @@ class Action:
                         if role == "user"
                         else "助手" if role == "assistant" else role
                     )
-                    aggregated_parts.append(f"[{role_label} 消息 {i}]\n{text_content}")
+                    aggregated_parts.append(f"{text_content}")
 
             if not aggregated_parts:
                 raise ValueError("无法获取有效的用户消息内容。")
@@ -1169,6 +1659,45 @@ class Action:
                         user_language,
                     )
 
+            # 检查输出模式
+            if self.valves.OUTPUT_MODE == "image":
+                # 图片模式：使用 JavaScript 渲染并嵌入为 Markdown 图片
+                chat_id = self._extract_chat_id(body, body.get("metadata"))
+                message_id = self._extract_message_id(body, body.get("metadata"))
+
+                await self._emit_status(
+                    __event_emitter__,
+                    "📊 信息图: 正在渲染图片...",
+                    False,
+                )
+
+                if __event_call__:
+                    js_code = self._generate_image_js_code(
+                        unique_id=unique_id,
+                        chat_id=chat_id,
+                        message_id=message_id,
+                        infographic_syntax=infographic_syntax,
+                    )
+
+                    await __event_call__(
+                        {
+                            "type": "execute",
+                            "data": {"code": js_code},
+                        }
+                    )
+
+                await self._emit_status(
+                    __event_emitter__, "✅ 信息图: 图片生成完成!", True
+                )
+                await self._emit_notification(
+                    __event_emitter__,
+                    f"📊 信息图图片已生成，{user_name}！",
+                    "success",
+                )
+                logger.info("信息图生成完成（图片模式）")
+                return body
+
+            # HTML 模式（默认）：嵌入为 HTML 块
             html_embed_tag = f"```html\n{final_html}\n```"
             body["messages"][-1]["content"] = f"{original_content}\n\n{html_embed_tag}"
 

plugins/actions/infographic/verify_generation.py

@@ -24,7 +24,7 @@ if not API_KEY or not BASE_URL:
     sys.exit(1)
 
 # =================================================================
-# Prompts (Extracted from 信息图.py)
+# Prompts (Extracted from infographic_cn.py)
 # =================================================================
 
 SYSTEM_PROMPT_INFOGRAPHIC_ASSISTANT = """

plugins/actions/smart-mind-map/README.md

@@ -1,6 +1,6 @@
 # Smart Mind Map - Mind Mapping Generation Plugin
 
-**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.8.0 | **License:** MIT
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 0.9.1 | **License:** MIT
 
 > **Important**: To ensure the maintainability and usability of all plugins, each plugin should be accompanied by clear and comprehensive documentation to ensure its functionality, configuration, and usage are well explained.
 
@@ -8,6 +8,25 @@ Smart Mind Map is a powerful OpenWebUI action plugin that intelligently analyzes
 
 ---
 
+## 🔥 What's New in v0.9.1
+
+**New Feature: Image Output Mode**
+
+- **Static Image Support**: Added `OUTPUT_MODE` configuration parameter.
+  - `html` (default): Interactive HTML mind map.
+  - `image`: Static SVG image embedded directly in Markdown (**No HTML code output**, cleaner chat history).
+- **Efficient Storage**: Image mode uploads SVG to `/api/v1/files`, avoiding huge base64 strings in chat history.
+- **Smart Features**: Auto-responsive width and automatic theme detection (light/dark) for generated images.
+
+| Feature | HTML Mode (Default) | Image Mode |
+| :--- | :--- | :--- |
+| **Output Format** | Interactive HTML Block | Static Markdown Image |
+| **Interactivity** | Zoom, Pan, Expand/Collapse | None (Static Image) |
+| **Chat History** | Contains HTML Code | Clean (Image URL only) |
+| **Storage** | Browser Rendering | `/api/v1/files` Upload |
+
+---
+
 ## Core Features
 
 - ✅ **Intelligent Text Analysis**: Automatically identifies core themes, key concepts, and hierarchical structures
@@ -20,6 +39,7 @@ Smart Mind Map is a powerful OpenWebUI action plugin that intelligently analyzes
 - ✅ **Real-time Rendering**: Renders mind maps directly in the chat interface without navigation
 - ✅ **Export Capabilities**: Supports PNG, SVG code, and Markdown source export
 - ✅ **Customizable Configuration**: Configurable LLM model, minimum text length, and other parameters
+- ✅ **Image Output Mode**: Generate static SVG images embedded directly in Markdown (**No HTML code output**, cleaner chat history)
 
 ---
 
@@ -39,7 +59,7 @@ Smart Mind Map is a powerful OpenWebUI action plugin that intelligently analyzes
 
 ### 1. Plugin Installation
 
-1. Download the `思维导图.py` file to your local computer
+1. Download the `smart_mind_map_cn.py` file to your local computer
 2. In OpenWebUI Admin Settings, find the "Plugins" section
 3. Select "Actions" type
 4. Upload the downloaded file
@@ -80,6 +100,7 @@ You can adjust the following parameters in the plugin's settings (Valves):
 | `MIN_TEXT_LENGTH` | `100` | Minimum text length (in characters) required for mind map analysis. Text that's too short cannot generate valid mind maps. |
 | `CLEAR_PREVIOUS_HTML` | `false` | Whether to clear previous plugin-generated HTML content when generating a new mind map. |
 | `MESSAGE_COUNT` | `1` | Number of recent messages to use for mind map generation (1-5). |
+| `OUTPUT_MODE` | `html` | Output mode: `html` for interactive HTML (default), or `image` to embed as static Markdown image. |
 
 ---
 
@@ -277,7 +298,37 @@ This plugin uses only OpenWebUI's built-in dependencies. **No additional package
 
 ## Changelog
 
-### v0.8.0 (Current Version)
+### v0.9.1
+
+**New Feature: Image Output Mode**
+
+- Added `OUTPUT_MODE` configuration parameter with two options:
+  - `html` (default): Interactive HTML mind map with full control panel
+  - `image`: Static SVG image embedded directly in Markdown (uploaded to `/api/v1/files`)
+- Image mode features:
+  - Auto-responsive width (adapts to chat container)
+  - Automatic theme detection (light/dark)
+  - Persistent storage via Chat API (survives page refresh)
+  - Efficient file storage (no huge base64 strings in chat history)
+
+**Improvements:**
+
+- Implemented robust Chat API update mechanism with retry logic
+- Fixed message persistence using both `messages[]` and `history.messages`
+- Added Event API for immediate frontend updates
+- Removed unnecessary `SVG_WIDTH` and `SVG_HEIGHT` parameters (now auto-calculated)
+
+**Technical Details:**
+
+- Image mode uses `__event_call__` to execute JavaScript in the browser
+- SVG is rendered offline, converted to Blob, and uploaded to OpenWebUI Files API
+- Updates chat message with `/api/v1/files/{id}/content` URL via OpenWebUI Backend-Controlled API flow
+
+### v0.8.2
+
+- Removed debug messages from output
+
+### v0.8.0 (Previous Version)
 
 **Major Features:**
 

plugins/actions/smart-mind-map/README_CN.md

@@ -1,6 +1,6 @@
 # 思维导图 - 思维导图生成插件
 
-**作者:** [Fu-Jie](https://github.com/Fu-Jie) | **版本:** 0.8.0 | **许可证:** MIT
+**作者:** [Fu-Jie](https://github.com/Fu-Jie) | **版本:** 0.9.1 | **许可证:** MIT
 
 > **重要提示**：为了确保所有插件的可维护性和易用性，每个插件都应附带清晰、完整的文档，以确保其功能、配置和使用方法得到充分说明。
 
@@ -8,6 +8,25 @@
 
 ---
 
+## 🔥 v0.9.1 更新亮点
+
+**新功能：图片输出模式**
+
+- **静态图片支持**：新增 `OUTPUT_MODE` 配置参数。
+  - `html`（默认）：交互式 HTML 思维导图。
+  - `image`：静态 SVG 图片直接嵌入 Markdown（**不输出 HTML 代码**，聊天记录更简洁）。
+- **高效存储**：图片模式将 SVG 上传至 `/api/v1/files`，避免聊天记录中出现超长 Base64 字符串。
+- **智能特性**：生成的图片支持自动响应式宽度和自动主题检测（亮色/暗色）。
+
+| 特性 | HTML 模式 (默认) | 图片模式 |
+| :--- | :--- | :--- |
+| **输出格式** | 交互式 HTML 代码块 | 静态 Markdown 图片 |
+| **交互性** | 缩放、拖拽、展开/折叠 | 无 (静态图片) |
+| **聊天记录** | 包含 HTML 代码 | 简洁 (仅图片链接) |
+| **存储方式** | 浏览器实时渲染 | `/api/v1/files` 上传 |
+
+---
+
 ## 核心特性
 
 - ✅ **智能文本分析**：自动识别文本的核心主题、关键概念和层次结构
@@ -20,6 +39,7 @@
 - ✅ **实时渲染**：在聊天界面中直接渲染思维导图，无需跳转
 - ✅ **导出功能**：支持 PNG、SVG 代码和 Markdown 源码导出
 - ✅ **自定义配置**：可配置 LLM 模型、最小文本长度等参数
+- ✅ **图片输出模式**：生成静态 SVG 图片直接嵌入 Markdown（**不输出 HTML 代码**，聊天记录更简洁）
 
 ---
 
@@ -39,7 +59,7 @@
 
 ### 1. 插件安装
 
-1. 下载 `思维导图.py` 文件到本地
+1. 下载 `smart_mind_map_cn.py` 文件到本地
 2. 在 OpenWebUI 管理员设置中找到"插件"（Plugins）部分
 3. 选择"动作"（Actions）类型
 4. 上传下载的文件
@@ -80,6 +100,7 @@
 | `MIN_TEXT_LENGTH` | `100` | 进行思维导图分析所需的最小文本长度（字符数）。文本过短将无法生成有效的导图。 |
 | `CLEAR_PREVIOUS_HTML` | `false` | 在生成新的思维导图时，是否清除之前由插件生成的 HTML 内容。 |
 | `MESSAGE_COUNT` | `1` | 用于生成思维导图的最近消息数量（1-5）。 |
+| `OUTPUT_MODE` | `html` | 输出模式：`html` 为交互式 HTML（默认），`image` 为嵌入静态 Markdown 图片。 |
 
 ---
 
@@ -277,7 +298,37 @@
 
 ## 更新日志
 
-### v0.8.0（当前版本）
+### v0.9.1
+
+**新功能：图片输出模式**
+
+- 新增 `OUTPUT_MODE` 配置参数，支持两种模式：
+  - `html`（默认）：交互式 HTML 思维导图，带完整控制面板
+  - `image`：静态 SVG 图片直接嵌入 Markdown（上传至 `/api/v1/files`）
+- 图片模式特性：
+  - 自动响应式宽度（适应聊天容器）
+  - 自动主题检测（亮色/暗色）
+  - 通过 Chat API 持久化存储（刷新页面后保留）
+  - 高效文件存储（聊天记录中无超长 Base64 字符串）
+
+**改进项：**
+
+- 实现健壮的 Chat API 更新机制，带重试逻辑
+- 修复消息持久化，同时更新 `messages[]` 和 `history.messages`
+- 添加 Event API 实现即时前端更新
+- 移除不必要的 `SVG_WIDTH` 和 `SVG_HEIGHT` 参数（现已自动计算）
+
+**技术细节：**
+
+- 图片模式使用 `__event_call__` 在浏览器中执行 JavaScript
+- SVG 离屏渲染，转换为 Blob，并上传至 OpenWebUI Files API
+- 通过 OpenWebUI Backend-Controlled API 流程更新聊天消息为 `/api/v1/files/{id}/content` URL
+
+### v0.8.2
+
+- 移除输出中的调试信息
+
+### v0.8.0 (Previous Version)
 
 **主要功能：**
 

plugins/actions/smart-mind-map/smart_mind_map.py

@@ -3,7 +3,8 @@ title: Smart Mind Map
 author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.8.0
+version: 0.9.1
+openwebui_id: 3094c59a-b4dd-4e0c-9449-15e2dd547dc4
 icon_url: data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9ImN1cnJlbnRDb2xvciIgc3Ryb2tlLXdpZHRoPSIyIiBzdHJva2UtbGluZWNhcD0icm91bmQiIHN0cm9rZS1saW5lam9pbj0icm91bmQiPjxyZWN0IHg9IjE2IiB5PSIxNiIgd2lkdGg9IjYiIGhlaWdodD0iNiIgcng9IjEiLz48cmVjdCB4PSIyIiB5PSIxNiIgd2lkdGg9IjYiIGhlaWdodD0iNiIgcng9IjEiLz48cmVjdCB4PSI5IiB5PSIyIiB3aWR0aD0iNiIgaGVpZ2h0PSI2IiByeD0iMSIvPjxwYXRoIGQ9Ik01IDE2di0zYTEgMSAwIDAgMSAxLTFoMTJhMSAxIDAgMCAxIDEgMXYzIi8+PHBhdGggZD0iTTEyIDEyVjgiLz48L3N2Zz4=
 description: Intelligently analyzes text content and generates interactive mind maps to help users structure and visualize knowledge.
 """
@@ -13,7 +14,7 @@ import os
 import re
 import time
 from datetime import datetime, timezone
-from typing import Any, Dict, Optional
+from typing import Any, Callable, Awaitable, Dict, Optional
 from zoneinfo import ZoneInfo
 
 from fastapi import Request
@@ -786,6 +787,10 @@ class Action:
             default=1,
             description="Number of recent messages to use for generation. Set to 1 for just the last message, or higher for more context.",
         )
+        OUTPUT_MODE: str = Field(
+            default="html",
+            description="Output mode: 'html' for interactive HTML (default), or 'image' to embed as Markdown image.",
+        )
 
     def __init__(self):
         self.valves = self.Valves()
@@ -814,6 +819,46 @@ class Action:
             "user_language": user_data.get("language", "en-US"),
         }
 
+    def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract chat_id from body or metadata"""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                chat_id = body_metadata.get("chat_id")
+                if isinstance(chat_id, str) and chat_id.strip():
+                    return chat_id.strip()
+
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+        return ""
+
+    def _extract_message_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract message_id from body or metadata"""
+        if isinstance(body, dict):
+            message_id = body.get("id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                message_id = body_metadata.get("message_id")
+                if isinstance(message_id, str) and message_id.strip():
+                    return message_id.strip()
+
+        if isinstance(metadata, dict):
+            message_id = metadata.get("message_id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+        return ""
+
     def _extract_markdown_syntax(self, llm_output: str) -> str:
         match = re.search(r"```markdown\s*(.*?)\s*```", llm_output, re.DOTALL)
         if match:
@@ -901,14 +946,391 @@ class Action:
 
         return base_html.strip()
 
+    def _generate_image_js_code(
+        self,
+        unique_id: str,
+        chat_id: str,
+        message_id: str,
+        markdown_syntax: str,
+    ) -> str:
+        """Generate JavaScript code for frontend SVG rendering and image embedding"""
+
+        # Escape the syntax for JS embedding
+        syntax_escaped = (
+            markdown_syntax.replace("\\", "\\\\")
+            .replace("`", "\\`")
+            .replace("${", "\\${")
+            .replace("</script>", "<\\/script>")
+        )
+
+        return f"""
+(async function() {{
+    const uniqueId = "{unique_id}";
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    const defaultWidth = 1200;
+    const defaultHeight = 800;
+    
+    // Theme detection - check parent document for OpenWebUI theme
+    const detectTheme = () => {{
+        try {{
+            // 1. Check parent document's html/body class or data-theme
+            const html = document.documentElement;
+            const body = document.body;
+            const htmlClass = html ? html.className : '';
+            const bodyClass = body ? body.className : '';
+            const htmlDataTheme = html ? html.getAttribute('data-theme') : '';
+            
+            if (htmlDataTheme === 'dark' || bodyClass.includes('dark') || htmlClass.includes('dark')) {{
+                return 'dark';
+            }}
+            if (htmlDataTheme === 'light' || bodyClass.includes('light') || htmlClass.includes('light')) {{
+                return 'light';
+            }}
+            
+            // 2. Check meta theme-color
+            const metas = document.querySelectorAll('meta[name="theme-color"]');
+            if (metas.length > 0) {{
+                const color = metas[metas.length - 1].content.trim();
+                const m = color.match(/^#?([0-9a-f]{{6}})$/i);
+                if (m) {{
+                    const hex = m[1];
+                    const r = parseInt(hex.slice(0, 2), 16);
+                    const g = parseInt(hex.slice(2, 4), 16);
+                    const b = parseInt(hex.slice(4, 6), 16);
+                    const luma = (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+                    return luma < 0.5 ? 'dark' : 'light';
+                }}
+            }}
+            
+            // 3. Check system preference
+            if (window.matchMedia && window.matchMedia('(prefers-color-scheme: dark)').matches) {{
+                return 'dark';
+            }}
+            
+            return 'light';
+        }} catch (e) {{
+            return 'light';
+        }}
+    }};
+    
+    const currentTheme = detectTheme();
+    console.log("[MindMap Image] Detected theme:", currentTheme);
+    
+    // Theme-based colors
+    const colors = currentTheme === 'dark' ? {{
+        background: '#1f2937',
+        text: '#e5e7eb',
+        link: '#94a3b8',
+        nodeStroke: '#64748b'
+    }} : {{
+        background: '#ffffff',
+        text: '#1f2937',
+        link: '#546e7a',
+        nodeStroke: '#94a3b8'
+    }};
+    
+    // Auto-detect chat container width for responsive sizing
+    let svgWidth = defaultWidth;
+    let svgHeight = defaultHeight;
+    const chatContainer = document.getElementById('chat-container');
+    if (chatContainer) {{
+        const containerWidth = chatContainer.clientWidth;
+        if (containerWidth > 100) {{
+            // Use container width with some padding (90% of container)
+            svgWidth = Math.floor(containerWidth * 0.9);
+            // Maintain aspect ratio based on default dimensions
+            svgHeight = Math.floor(svgWidth * (defaultHeight / defaultWidth));
+            console.log("[MindMap Image] Auto-detected container width:", containerWidth, "-> SVG:", svgWidth, "x", svgHeight);
+        }}
+    }}
+    
+    console.log("[MindMap Image] Starting render...");
+    console.log("[MindMap Image] chatId:", chatId, "messageId:", messageId);
+    
+    try {{
+        // Load D3 if not loaded
+        if (typeof d3 === 'undefined') {{
+            console.log("[MindMap Image] Loading D3...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/d3@7';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        // Load markmap-lib if not loaded
+        if (!window.markmap || !window.markmap.Transformer) {{
+            console.log("[MindMap Image] Loading markmap-lib...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/markmap-lib@0.17';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        // Load markmap-view if not loaded
+        if (!window.markmap || !window.markmap.Markmap) {{
+            console.log("[MindMap Image] Loading markmap-view...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/markmap-view@0.17';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        const {{ Transformer, Markmap }} = window.markmap;
+        
+        // Get markdown syntax
+        let syntaxContent = `{syntax_escaped}`;
+        console.log("[MindMap Image] Syntax length:", syntaxContent.length);
+        
+        // Create offscreen container
+        const container = document.createElement('div');
+        container.id = 'mindmap-offscreen-' + uniqueId;
+        container.style.cssText = 'position:absolute;left:-9999px;top:-9999px;width:' + svgWidth + 'px;height:' + svgHeight + 'px;';
+        document.body.appendChild(container);
+        
+        // Create SVG element
+        const svgEl = document.createElementNS('http://www.w3.org/2000/svg', 'svg');
+        svgEl.setAttribute('width', svgWidth);
+        svgEl.setAttribute('height', svgHeight);
+        svgEl.style.width = svgWidth + 'px';
+        svgEl.style.height = svgHeight + 'px';
+        svgEl.style.backgroundColor = colors.background;
+        container.appendChild(svgEl);
+        
+        // Transform markdown to tree
+        const transformer = new Transformer();
+        const {{ root }} = transformer.transform(syntaxContent);
+        
+        // Create markmap instance
+        const options = {{
+            autoFit: true,
+            initialExpandLevel: Infinity,
+            zoom: false,
+            pan: false
+        }};
+        
+        console.log("[MindMap Image] Rendering markmap...");
+        const markmapInstance = Markmap.create(svgEl, options, root);
+        
+        // Wait for render to complete
+        await new Promise(resolve => setTimeout(resolve, 1500));
+        markmapInstance.fit();
+        await new Promise(resolve => setTimeout(resolve, 500));
+        
+        // Clone and prepare SVG for export
+        const clonedSvg = svgEl.cloneNode(true);
+        clonedSvg.setAttribute('xmlns', 'http://www.w3.org/2000/svg');
+        clonedSvg.setAttribute('xmlns:xlink', 'http://www.w3.org/1999/xlink');
+        
+        // Add background rect with theme color
+        const bgRect = document.createElementNS('http://www.w3.org/2000/svg', 'rect');
+        bgRect.setAttribute('width', '100%');
+        bgRect.setAttribute('height', '100%');
+        bgRect.setAttribute('fill', colors.background);
+        clonedSvg.insertBefore(bgRect, clonedSvg.firstChild);
+        
+        // Add inline styles with theme colors
+        const style = document.createElementNS('http://www.w3.org/2000/svg', 'style');
+        style.textContent = `
+            text {{ font-family: sans-serif; font-size: 14px; fill: ${{colors.text}}; }}
+            foreignObject, .markmap-foreign, .markmap-foreign div {{ color: ${{colors.text}}; font-family: sans-serif; font-size: 14px; }}
+            h1 {{ font-size: 22px; font-weight: 700; margin: 0; }}
+            h2 {{ font-size: 18px; font-weight: 600; margin: 0; }}
+            strong {{ font-weight: 700; }}
+            .markmap-link {{ stroke: ${{colors.link}}; fill: none; }}
+            .markmap-node circle, .markmap-node rect {{ stroke: ${{colors.nodeStroke}}; }}
+        `;
+        clonedSvg.insertBefore(style, bgRect.nextSibling);
+        
+        // Convert foreignObject to text for better compatibility
+        const foreignObjects = clonedSvg.querySelectorAll('foreignObject');
+        foreignObjects.forEach(fo => {{
+            const text = fo.textContent || '';
+            const g = document.createElementNS('http://www.w3.org/2000/svg', 'g');
+            const textEl = document.createElementNS('http://www.w3.org/2000/svg', 'text');
+            textEl.setAttribute('x', fo.getAttribute('x') || '0');
+            textEl.setAttribute('y', (parseFloat(fo.getAttribute('y') || '0') + 14).toString());
+            textEl.setAttribute('fill', colors.text);
+            textEl.setAttribute('font-family', 'sans-serif');
+            textEl.setAttribute('font-size', '14');
+            textEl.textContent = text.trim();
+            g.appendChild(textEl);
+            fo.parentNode.replaceChild(g, fo);
+        }});
+        
+        // Serialize SVG to string
+        const svgData = new XMLSerializer().serializeToString(clonedSvg);
+        
+        // Cleanup container
+        document.body.removeChild(container);
+        
+        // Convert SVG string to Blob
+        const blob = new Blob([svgData], {{ type: 'image/svg+xml' }});
+        const file = new File([blob], `mindmap-${{uniqueId}}.svg`, {{ type: 'image/svg+xml' }});
+        
+        // Upload file to OpenWebUI API
+        console.log("[MindMap Image] Uploading SVG file...");
+        const token = localStorage.getItem("token");
+        const formData = new FormData();
+        formData.append('file', file);
+        
+        const uploadResponse = await fetch('/api/v1/files/', {{
+            method: 'POST',
+            headers: {{
+                'Authorization': `Bearer ${{token}}`
+            }},
+            body: formData
+        }});
+        
+        if (!uploadResponse.ok) {{
+            throw new Error(`Upload failed: ${{uploadResponse.statusText}}`);
+        }}
+        
+        const fileData = await uploadResponse.json();
+        const fileId = fileData.id;
+        const imageUrl = `/api/v1/files/${{fileId}}/content`;
+        
+        console.log("[MindMap Image] File uploaded, ID:", fileId);
+        
+        // Generate markdown image with file URL
+        const markdownImage = `![🧠 Mind Map](${{imageUrl}})`;
+        
+        // Update message via API
+        if (chatId && messageId) {{
+            
+            // Helper function with retry logic
+            const fetchWithRetry = async (url, options, retries = 3) => {{
+                for (let i = 0; i < retries; i++) {{
+                    try {{
+                        const response = await fetch(url, options);
+                        if (response.ok) return response;
+                        if (i < retries - 1) {{
+                            console.log(`[MindMap Image] Retry ${{i + 1}}/${{retries}} for ${{url}}`);
+                            await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                        }}
+                    }} catch (e) {{
+                        if (i === retries - 1) throw e;
+                        await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                    }}
+                }}
+                return null;
+            }};
+            
+            // Get current chat data
+            const getResponse = await fetch(`/api/v1/chats/${{chatId}}`, {{
+                method: "GET",
+                headers: {{ "Authorization": `Bearer ${{token}}` }}
+            }});
+            
+            if (!getResponse.ok) {{
+                throw new Error("Failed to get chat data: " + getResponse.status);
+            }}
+            
+            const chatData = await getResponse.json();
+            let updatedMessages = [];
+            let newContent = "";
+            
+            if (chatData.chat && chatData.chat.messages) {{
+                updatedMessages = chatData.chat.messages.map(m => {{
+                    if (m.id === messageId) {{
+                        const originalContent = m.content || "";
+                        // Remove existing mindmap images (both base64 and file URL patterns)
+                        const mindmapPattern = /\\n*!\\[🧠[^\\]]*\\]\\((?:data:image\\/[^)]+|(?:\\/api\\/v1\\/files\\/[^)]+))\\)/g;
+                        let cleanedContent = originalContent.replace(mindmapPattern, "");
+                        cleanedContent = cleanedContent.replace(/\\n{{3,}}/g, "\\n\\n").trim();
+                        // Append new image
+                        newContent = cleanedContent + "\\n\\n" + markdownImage;
+                        
+                        // Critical: Update content in both messages array AND history object
+                        // The history object is the source of truth for the database
+                        if (chatData.chat.history && chatData.chat.history.messages) {{
+                            if (chatData.chat.history.messages[messageId]) {{
+                                chatData.chat.history.messages[messageId].content = newContent;
+                            }}
+                        }}
+                        
+                        return {{ ...m, content: newContent }};
+                    }}
+                    return m;
+                }});
+            }}
+            
+            if (!newContent) {{
+                console.warn("[MindMap Image] Could not find message to update");
+                return;
+            }}
+            
+            // Try to update frontend display via event API (optional, may not exist in all versions)
+            try {{
+                await fetch(`/api/v1/chats/${{chatId}}/messages/${{messageId}}/event`, {{
+                    method: "POST",
+                    headers: {{
+                        "Content-Type": "application/json",
+                        "Authorization": `Bearer ${{token}}`
+                    }},
+                    body: JSON.stringify({{
+                        type: "chat:message",
+                        data: {{ content: newContent }}
+                    }})
+                }});
+            }} catch (eventErr) {{
+                // Event API is optional, continue with persistence
+                console.log("[MindMap Image] Event API not available, continuing...");
+            }}
+            
+            // Persist to database by updating the entire chat object
+            // This follows the OpenWebUI Backend-Controlled API Flow
+            const updatePayload = {{
+                chat: {{
+                    ...chatData.chat,
+                    messages: updatedMessages
+                    // history is already updated in-place above
+                }}
+            }};
+            
+            const persistResponse = await fetchWithRetry(`/api/v1/chats/${{chatId}}`, {{
+                method: "POST",
+                headers: {{
+                    "Content-Type": "application/json",
+                    "Authorization": `Bearer ${{token}}`
+                }},
+                body: JSON.stringify(updatePayload)
+            }});
+            
+            if (persistResponse && persistResponse.ok) {{
+                console.log("[MindMap Image] ✅ Message persisted successfully!");
+            }} else {{
+                console.error("[MindMap Image] ❌ Failed to persist message after retries");
+            }}
+        }} else {{
+            console.warn("[MindMap Image] ⚠️ Missing chatId or messageId, cannot persist");
+        }}
+        
+    }} catch (error) {{
+        console.error("[MindMap Image] Error:", error);
+    }}
+}})();
+"""
+
     async def action(
         self,
         body: dict,
         __user__: Optional[Dict[str, Any]] = None,
         __event_emitter__: Optional[Any] = None,
+        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __metadata__: Optional[dict] = None,
         __request__: Optional[Request] = None,
     ) -> Optional[dict]:
-        logger.info("Action: Smart Mind Map (v0.8.0) started")
+        logger.info("Action: Smart Mind Map (v0.9.1) started")
         user_ctx = self._get_user_context(__user__)
         user_language = user_ctx["user_language"]
         user_name = user_ctx["user_name"]
@@ -960,7 +1382,7 @@ class Action:
                     if role == "user"
                     else "Assistant" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} Message {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             error_message = "Unable to retrieve valid user message content."
@@ -1090,6 +1512,45 @@ class Action:
                         user_language,
                     )
 
+            # Check output mode
+            if self.valves.OUTPUT_MODE == "image":
+                # Image mode: use JavaScript to render and embed as Markdown image
+                chat_id = self._extract_chat_id(body, __metadata__)
+                message_id = self._extract_message_id(body, __metadata__)
+
+                await self._emit_status(
+                    __event_emitter__,
+                    "Smart Mind Map: Rendering image...",
+                    False,
+                )
+
+                if __event_call__:
+                    js_code = self._generate_image_js_code(
+                        unique_id=unique_id,
+                        chat_id=chat_id,
+                        message_id=message_id,
+                        markdown_syntax=markdown_syntax,
+                    )
+
+                    await __event_call__(
+                        {
+                            "type": "execute",
+                            "data": {"code": js_code},
+                        }
+                    )
+
+                await self._emit_status(
+                    __event_emitter__, "Smart Mind Map: Image generated!", True
+                )
+                await self._emit_notification(
+                    __event_emitter__,
+                    f"Mind map image has been generated, {user_name}!",
+                    "success",
+                )
+                logger.info("Action: Smart Mind Map (v0.9.1) completed in image mode")
+                return body
+
+            # HTML mode (default): embed as HTML block
             html_embed_tag = f"```html\n{final_html}\n```"
             body["messages"][-1]["content"] = f"{long_text_content}\n\n{html_embed_tag}"
 
@@ -1101,7 +1562,7 @@ class Action:
                 f"Mind map has been generated, {user_name}!",
                 "success",
             )
-            logger.info("Action: Smart Mind Map (v0.8.0) completed successfully")
+            logger.info("Action: Smart Mind Map (v0.9.1) completed in HTML mode")
 
         except Exception as e:
             error_message = f"Smart Mind Map processing failed: {str(e)}"

plugins/actions/smart-mind-map/smart_mind_map_cn.py

@@ -3,7 +3,8 @@ title: 思维导图
 author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.8.0
+version: 0.9.1
+openwebui_id: 8d4b097b-219b-4dd2-b509-05fbe6388335
 icon_url: data:image/svg+xml;base64,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
 description: 智能分析文本内容,生成交互式思维导图,帮助用户结构化和可视化知识。
 """
@@ -13,7 +14,7 @@ import os
 import re
 import time
 from datetime import datetime, timezone
-from typing import Any, Dict, Optional
+from typing import Any, Callable, Awaitable, Dict, Optional
 from zoneinfo import ZoneInfo
 
 from fastapi import Request
@@ -443,7 +444,7 @@ SCRIPT_TEMPLATE_MINDMAP = """
 
             const markdownContent = sourceEl.textContent.trim();
             if (!markdownContent) {
-                containerEl.innerHTML = '<div class="error-message">⚠️ 无法加载思维导图：缺少有效内容。</div>';
+                containerEl.innerHTML = '<div class="error-message">⚠️ 无法加载思维导图:缺少有效内容。</div>';
                 return;
             }
 
@@ -485,7 +486,7 @@ SCRIPT_TEMPLATE_MINDMAP = """
 
             }).catch((error) => {
                 console.error('Markmap loading error:', error);
-                containerEl.innerHTML = '<div class="error-message">⚠️ 资源加载失败，请稍后重试。</div>';
+                containerEl.innerHTML = '<div class="error-message">⚠️ 资源加载失败,请稍后重试。</div>';
             });
         };
 
@@ -771,19 +772,23 @@ class Action:
         )
         MODEL_ID: str = Field(
             default="",
-            description="用于文本分析的内置LLM模型ID。如果为空，则使用当前对话的模型。",
+            description="用于文本分析的内置LLM模型ID。如果为空,则使用当前对话的模型。",
         )
         MIN_TEXT_LENGTH: int = Field(
             default=100,
-            description="进行思维导图分析所需的最小文本长度（字符数）。",
+            description="进行思维导图分析所需的最小文本长度(字符数)。",
         )
         CLEAR_PREVIOUS_HTML: bool = Field(
             default=False,
-            description="是否强制清除旧的插件结果（如果为 True，则不合并，直接覆盖）。",
+            description="是否强制清除旧的插件结果(如果为 True,则不合并,直接覆盖)。",
         )
         MESSAGE_COUNT: int = Field(
             default=1,
-            description="用于生成的最近消息数量。设置为1仅使用最后一条消息，更大值可包含更多上下文。",
+            description="用于生成的最近消息数量。设置为1仅使用最后一条消息,更大值可包含更多上下文。",
+        )
+        OUTPUT_MODE: str = Field(
+            default="html",
+            description="输出模式: 'html' 为交互式HTML(默认),'image' 为嵌入Markdown图片。",
         )
 
     def __init__(self):
@@ -813,14 +818,52 @@ class Action:
             "user_language": user_data.get("language", "zh-CN"),
         }
 
+    def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """从 body 或 metadata 中提取 chat_id"""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                chat_id = body_metadata.get("chat_id")
+                if isinstance(chat_id, str) and chat_id.strip():
+                    return chat_id.strip()
+
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+        return ""
+
+    def _extract_message_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """从 body 或 metadata 中提取 message_id"""
+        if isinstance(body, dict):
+            message_id = body.get("id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                message_id = body_metadata.get("message_id")
+                if isinstance(message_id, str) and message_id.strip():
+                    return message_id.strip()
+
+        if isinstance(metadata, dict):
+            message_id = metadata.get("message_id")
+            if isinstance(message_id, str) and message_id.strip():
+                return message_id.strip()
+
+        return ""
+
     def _extract_markdown_syntax(self, llm_output: str) -> str:
         match = re.search(r"```markdown\s*(.*?)\s*```", llm_output, re.DOTALL)
         if match:
             extracted_content = match.group(1).strip()
         else:
-            logger.warning(
-                "LLM输出未严格遵循预期Markdown格式，将整个输出作为摘要处理。"
-            )
+            logger.warning("LLM输出未严格遵循预期Markdown格式,将整个输出作为摘要处理。")
             extracted_content = llm_output.strip()
         return extracted_content.replace("</script>", "<\\/script>")
 
@@ -844,7 +887,7 @@ class Action:
         return re.sub(pattern, "", content).strip()
 
     def _extract_text_content(self, content) -> str:
-        """从消息内容中提取文本，支持多模态消息格式"""
+        """从消息内容中提取文本,支持多模态消息格式"""
         if isinstance(content, str):
             return content
         elif isinstance(content, list):
@@ -867,7 +910,7 @@ class Action:
         user_language: str = "zh-CN",
     ) -> str:
         """
-        将新内容合并到现有的 HTML 容器中，或者创建一个新的容器。
+        将新内容合并到现有的 HTML 容器中,或者创建一个新的容器。
         """
         if (
             "<!-- OPENWEBUI_PLUGIN_OUTPUT -->" in existing_html_code
@@ -900,14 +943,392 @@ class Action:
 
         return base_html.strip()
 
+    def _generate_image_js_code(
+        self,
+        unique_id: str,
+        chat_id: str,
+        message_id: str,
+        markdown_syntax: str,
+    ) -> str:
+        """生成用于前端 SVG 渲染和图片嵌入的 JavaScript 代码"""
+
+        # 转义语法以便嵌入 JS
+        syntax_escaped = (
+            markdown_syntax.replace("\\", "\\\\")
+            .replace("`", "\\`")
+            .replace("${", "\\${")
+            .replace("</script>", "<\\/script>")
+        )
+
+        return f"""
+(async function() {{
+    const uniqueId = "{unique_id}";
+    const chatId = "{chat_id}";
+    const messageId = "{message_id}";
+    const defaultWidth = 1200;
+    const defaultHeight = 800;
+    
+    // 主题检测 - 检查 OpenWebUI 当前主题
+    const detectTheme = () => {{
+        try {{
+            // 1. 检查 html/body 的 class 或 data-theme 属性
+            const html = document.documentElement;
+            const body = document.body;
+            const htmlClass = html ? html.className : '';
+            const bodyClass = body ? body.className : '';
+            const htmlDataTheme = html ? html.getAttribute('data-theme') : '';
+            
+            if (htmlDataTheme === 'dark' || bodyClass.includes('dark') || htmlClass.includes('dark')) {{
+                return 'dark';
+            }}
+            if (htmlDataTheme === 'light' || bodyClass.includes('light') || htmlClass.includes('light')) {{
+                return 'light';
+            }}
+            
+            // 2. 检查 meta theme-color
+            const metas = document.querySelectorAll('meta[name="theme-color"]');
+            if (metas.length > 0) {{
+                const color = metas[metas.length - 1].content.trim();
+                const m = color.match(/^#?([0-9a-f]{{6}})$/i);
+                if (m) {{
+                    const hex = m[1];
+                    const r = parseInt(hex.slice(0, 2), 16);
+                    const g = parseInt(hex.slice(2, 4), 16);
+                    const b = parseInt(hex.slice(4, 6), 16);
+                    const luma = (0.2126 * r + 0.7152 * g + 0.0722 * b) / 255;
+                    return luma < 0.5 ? 'dark' : 'light';
+                }}
+            }}
+            
+            // 3. 检查系统偏好
+            if (window.matchMedia && window.matchMedia('(prefers-color-scheme: dark)').matches) {{
+                return 'dark';
+            }}
+            
+            return 'light';
+        }} catch (e) {{
+            return 'light';
+        }}
+    }};
+    
+    const currentTheme = detectTheme();
+    console.log("[思维导图图片] 检测到主题:", currentTheme);
+    
+    // 基于主题的颜色配置
+    const colors = currentTheme === 'dark' ? {{
+        background: '#1f2937',
+        text: '#e5e7eb',
+        link: '#94a3b8',
+        nodeStroke: '#64748b'
+    }} : {{
+        background: '#ffffff',
+        text: '#1f2937',
+        link: '#546e7a',
+        nodeStroke: '#94a3b8'
+    }};
+    
+    // 自动检测聊天容器宽度以实现自适应
+    let svgWidth = defaultWidth;
+    let svgHeight = defaultHeight;
+    const chatContainer = document.getElementById('chat-container');
+    if (chatContainer) {{
+        const containerWidth = chatContainer.clientWidth;
+        if (containerWidth > 100) {{
+            // 使用容器宽度的90%(留出边距)
+            svgWidth = Math.floor(containerWidth * 0.9);
+            // 根据默认尺寸保持宽高比
+            svgHeight = Math.floor(svgWidth * (defaultHeight / defaultWidth));
+            console.log("[思维导图图片] 自动检测容器宽度:", containerWidth, "-> SVG:", svgWidth, "x", svgHeight);
+        }}
+    }}
+    
+    console.log("[思维导图图片] 开始渲染...");
+    console.log("[思维导图图片] chatId:", chatId, "messageId:", messageId);
+    
+    try {{
+        // 加载 D3
+        if (typeof d3 === 'undefined') {{
+            console.log("[思维导图图片] 正在加载 D3...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/d3@7';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        // 加载 markmap-lib
+        if (!window.markmap || !window.markmap.Transformer) {{
+            console.log("[思维导图图片] 正在加载 markmap-lib...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/markmap-lib@0.17';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        // 加载 markmap-view
+        if (!window.markmap || !window.markmap.Markmap) {{
+            console.log("[思维导图图片] 正在加载 markmap-view...");
+            await new Promise((resolve, reject) => {{
+                const script = document.createElement('script');
+                script.src = 'https://cdn.jsdelivr.net/npm/markmap-view@0.17';
+                script.onload = resolve;
+                script.onerror = reject;
+                document.head.appendChild(script);
+            }});
+        }}
+        
+        const {{ Transformer, Markmap }} = window.markmap;
+        
+        // 获取 markdown 语法
+        let syntaxContent = `{syntax_escaped}`;
+        console.log("[思维导图图片] 语法长度:", syntaxContent.length);
+        
+        // 创建离屏容器
+        const container = document.createElement('div');
+        container.id = 'mindmap-offscreen-' + uniqueId;
+        container.style.cssText = 'position:absolute;left:-9999px;top:-9999px;width:' + svgWidth + 'px;height:' + svgHeight + 'px;';
+        document.body.appendChild(container);
+        
+        // 创建 SVG 元素
+        const svgEl = document.createElementNS('http://www.w3.org/2000/svg', 'svg');
+        svgEl.setAttribute('width', svgWidth);
+        svgEl.setAttribute('height', svgHeight);
+        svgEl.style.width = svgWidth + 'px';
+        svgEl.style.height = svgHeight + 'px';
+        svgEl.style.backgroundColor = colors.background;
+        container.appendChild(svgEl);
+        
+        // 将 markdown 转换为树结构
+        const transformer = new Transformer();
+        const {{ root }} = transformer.transform(syntaxContent);
+        
+        // 创建 markmap 实例
+        const options = {{
+            autoFit: true,
+            initialExpandLevel: Infinity,
+            zoom: false,
+            pan: false
+        }};
+        
+        console.log("[思维导图图片] 正在渲染 markmap...");
+        const markmapInstance = Markmap.create(svgEl, options, root);
+        
+        // 等待渲染完成
+        await new Promise(resolve => setTimeout(resolve, 1500));
+        markmapInstance.fit();
+        await new Promise(resolve => setTimeout(resolve, 500));
+        
+        // 克隆并准备 SVG 导出
+        const clonedSvg = svgEl.cloneNode(true);
+        clonedSvg.setAttribute('xmlns', 'http://www.w3.org/2000/svg');
+        clonedSvg.setAttribute('xmlns:xlink', 'http://www.w3.org/1999/xlink');
+        
+        // 添加背景矩形(使用主题颜色)
+        const bgRect = document.createElementNS('http://www.w3.org/2000/svg', 'rect');
+        bgRect.setAttribute('width', '100%');
+        bgRect.setAttribute('height', '100%');
+        bgRect.setAttribute('fill', colors.background);
+        clonedSvg.insertBefore(bgRect, clonedSvg.firstChild);
+        
+        // 添加内联样式(使用主题颜色)
+        const style = document.createElementNS('http://www.w3.org/2000/svg', 'style');
+        style.textContent = `
+            text {{ font-family: sans-serif; font-size: 14px; fill: ${{colors.text}}; }}
+            foreignObject, .markmap-foreign, .markmap-foreign div {{ color: ${{colors.text}}; font-family: sans-serif; font-size: 14px; }}
+            h1 {{ font-size: 22px; font-weight: 700; margin: 0; }}
+            h2 {{ font-size: 18px; font-weight: 600; margin: 0; }}
+            strong {{ font-weight: 700; }}
+            .markmap-link {{ stroke: ${{colors.link}}; fill: none; }}
+            .markmap-node circle, .markmap-node rect {{ stroke: ${{colors.nodeStroke}}; }}
+        `;
+        clonedSvg.insertBefore(style, bgRect.nextSibling);
+        
+        // 将 foreignObject 转换为 text 以提高兼容性
+        const foreignObjects = clonedSvg.querySelectorAll('foreignObject');
+        foreignObjects.forEach(fo => {{
+            const text = fo.textContent || '';
+            const g = document.createElementNS('http://www.w3.org/2000/svg', 'g');
+            const textEl = document.createElementNS('http://www.w3.org/2000/svg', 'text');
+            textEl.setAttribute('x', fo.getAttribute('x') || '0');
+            textEl.setAttribute('y', (parseFloat(fo.getAttribute('y') || '0') + 14).toString());
+            textEl.setAttribute('fill', colors.text);
+            textEl.setAttribute('font-family', 'sans-serif');
+            textEl.setAttribute('font-size', '14');
+            textEl.textContent = text.trim();
+            g.appendChild(textEl);
+            fo.parentNode.replaceChild(g, fo);
+        }});
+        
+        // 序列化 SVG 为字符串
+        const svgData = new XMLSerializer().serializeToString(clonedSvg);
+        
+        // 清理容器
+        document.body.removeChild(container);
+        
+        // 将 SVG 字符串转换为 Blob
+        const blob = new Blob([svgData], {{ type: 'image/svg+xml' }});
+        const file = new File([blob], `mindmap-${{uniqueId}}.svg`, {{ type: 'image/svg+xml' }});
+        
+        // 上传文件到 OpenWebUI API
+        console.log("[思维导图图片] 正在上传 SVG 文件...");
+        const token = localStorage.getItem("token");
+        const formData = new FormData();
+        formData.append('file', file);
+        
+        const uploadResponse = await fetch('/api/v1/files/', {{
+            method: 'POST',
+            headers: {{
+                'Authorization': `Bearer ${{token}}`
+            }},
+            body: formData
+        }});
+        
+        if (!uploadResponse.ok) {{
+            throw new Error(`上传失败: ${{uploadResponse.statusText}}`);
+        }}
+        
+        const fileData = await uploadResponse.json();
+        const fileId = fileData.id;
+        const imageUrl = `/api/v1/files/${{fileId}}/content`;
+        
+        console.log("[思维导图图片] 文件已上传, ID:", fileId);
+        
+        // 生成包含文件 URL 的 markdown 图片
+        const markdownImage = `![🧠 思维导图](${{imageUrl}})`;
+        
+        // 通过 API 更新消息
+        if (chatId && messageId) {{
+            const token = localStorage.getItem("token");
+            
+            // 带重试逻辑的请求函数
+            const fetchWithRetry = async (url, options, retries = 3) => {{
+                for (let i = 0; i < retries; i++) {{
+                    try {{
+                        const response = await fetch(url, options);
+                        if (response.ok) return response;
+                        if (i < retries - 1) {{
+                            console.log(`[思维导图图片] 重试 ${{i + 1}}/${{retries}}: ${{url}}`);
+                            await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                        }}
+                    }} catch (e) {{
+                        if (i === retries - 1) throw e;
+                        await new Promise(r => setTimeout(r, 1000 * (i + 1)));
+                    }}
+                }}
+                return null;
+            }};
+            
+            // 获取当前聊天数据
+            const getResponse = await fetch(`/api/v1/chats/${{chatId}}`, {{
+                method: "GET",
+                headers: {{ "Authorization": `Bearer ${{token}}` }}
+            }});
+            
+            if (!getResponse.ok) {{
+                throw new Error("获取聊天数据失败: " + getResponse.status);
+            }}
+            
+            const chatData = await getResponse.json();
+            let updatedMessages = [];
+            let newContent = "";
+            
+            if (chatData.chat && chatData.chat.messages) {{
+                updatedMessages = chatData.chat.messages.map(m => {{
+                    if (m.id === messageId) {{
+                        const originalContent = m.content || "";
+                        // 移除已有的思维导图图片 (包括 base64 和文件 URL 格式)
+                        const mindmapPattern = /\\n*!\\[🧠[^\\]]*\\]\\((?:data:image\\/[^)]+|(?:\\/api\\/v1\\/files\\/[^)]+))\\)/g;
+                        let cleanedContent = originalContent.replace(mindmapPattern, "");
+                        cleanedContent = cleanedContent.replace(/\\n{{3,}}/g, "\\n\\n").trim();
+                        // 追加新图片
+                        newContent = cleanedContent + "\\n\\n" + markdownImage;
+                        
+                        // 关键: 同时更新 messages 数组和 history 对象中的内容
+                        // history 对象是数据库的单一真值来源
+                        if (chatData.chat.history && chatData.chat.history.messages) {{
+                            if (chatData.chat.history.messages[messageId]) {{
+                                chatData.chat.history.messages[messageId].content = newContent;
+                            }}
+                        }}
+                        
+                        return {{ ...m, content: newContent }};
+                    }}
+                    return m;
+                }});
+            }}
+            
+            if (!newContent) {{
+                console.warn("[思维导图图片] 找不到要更新的消息");
+                return;
+            }}
+            
+            // 尝试通过事件 API 更新前端显示(可选,部分版本可能不支持)
+            try {{
+                await fetch(`/api/v1/chats/${{chatId}}/messages/${{messageId}}/event`, {{
+                    method: "POST",
+                    headers: {{
+                        "Content-Type": "application/json",
+                        "Authorization": `Bearer ${{token}}`
+                    }},
+                    body: JSON.stringify({{
+                        type: "chat:message",
+                        data: {{ content: newContent }}
+                    }})
+                }});
+            }} catch (eventErr) {{
+                // 事件 API 是可选的,继续执行持久化
+                console.log("[思维导图图片] 事件 API 不可用,继续执行...");
+            }}
+            
+            // 通过更新整个聊天对象来持久化到数据库
+            // 遵循 OpenWebUI 后端控制的 API 流程
+            const updatePayload = {{
+                chat: {{
+                    ...chatData.chat,
+                    messages: updatedMessages
+                    // history 已在上面原地更新
+                }}
+            }};
+            
+            const persistResponse = await fetchWithRetry(`/api/v1/chats/${{chatId}}`, {{
+                method: "POST",
+                headers: {{
+                    "Content-Type": "application/json",
+                    "Authorization": `Bearer ${{token}}`
+                }},
+                body: JSON.stringify(updatePayload)
+            }});
+            
+            if (persistResponse && persistResponse.ok) {{
+                console.log("[思维导图图片] ✅ 消息已持久化保存!");
+            }} else {{
+                console.error("[思维导图图片] ❌ 重试后仍然无法持久化消息");
+            }}
+        }} else {{
+            console.warn("[思维导图图片] ⚠️ 缺少 chatId 或 messageId,无法持久化");
+        }}
+        
+    }} catch (error) {{
+        console.error("[思维导图图片] 错误:", error);
+    }}
+}})();
+"""
+
     async def action(
         self,
         body: dict,
         __user__: Optional[Dict[str, Any]] = None,
         __event_emitter__: Optional[Any] = None,
+        __event_call__: Optional[Callable[[Any], Awaitable[None]]] = None,
+        __metadata__: Optional[dict] = None,
         __request__: Optional[Request] = None,
     ) -> Optional[dict]:
-        logger.info("Action: 思维导图 (v12 - Final Feedback Fix) started")
+        logger.info("Action: 思维导图 (v0.9.1) started")
         user_ctx = self._get_user_context(__user__)
         user_language = user_ctx["user_language"]
         user_name = user_ctx["user_name"]
@@ -923,7 +1344,7 @@ class Action:
             current_year = now_dt.strftime("%Y")
             current_timezone_str = tz_env or "UTC"
         except Exception as e:
-            logger.warning(f"获取时区信息失败: {e}，使用默认值。")
+            logger.warning(f"获取时区信息失败: {e},使用默认值。")
             now = datetime.now()
             current_date_time_str = now.strftime("%Y年%m月%d日 %H:%M:%S")
             current_weekday_zh = "未知星期"
@@ -931,7 +1352,7 @@ class Action:
             current_timezone_str = "未知时区"
 
         await self._emit_notification(
-            __event_emitter__, "思维导图已启动，正在为您生成思维导图...", "info"
+            __event_emitter__, "思维导图已启动,正在为您生成思维导图...", "info"
         )
 
         messages = body.get("messages")
@@ -957,7 +1378,7 @@ class Action:
                     if role == "user"
                     else "助手" if role == "assistant" else role
                 )
-                aggregated_parts.append(f"[{role_label} 消息 {i}]\n{text_content}")
+                aggregated_parts.append(f"{text_content}")
 
         if not aggregated_parts:
             error_message = "无法获取有效的用户消息内容。"
@@ -980,7 +1401,7 @@ class Action:
             long_text_content = original_content.strip()
 
         if len(long_text_content) < self.valves.MIN_TEXT_LENGTH:
-            short_text_message = f"文本内容过短({len(long_text_content)}字符)，无法进行有效分析。请提供至少{self.valves.MIN_TEXT_LENGTH}字符的文本。"
+            short_text_message = f"文本内容过短({len(long_text_content)}字符),无法进行有效分析。请提供至少{self.valves.MIN_TEXT_LENGTH}字符的文本。"
             await self._emit_notification(
                 __event_emitter__, short_text_message, "warning"
             )
@@ -1021,7 +1442,7 @@ class Action:
             }
             user_obj = Users.get_user_by_id(user_id)
             if not user_obj:
-                raise ValueError(f"无法获取用户对象，用户ID: {user_id}")
+                raise ValueError(f"无法获取用户对象,用户ID: {user_id}")
 
             llm_response = await generate_chat_completion(
                 __request__, llm_payload, user_obj
@@ -1084,26 +1505,65 @@ class Action:
                         user_language,
                     )
 
+            # 检查输出模式
+            if self.valves.OUTPUT_MODE == "image":
+                # 图片模式: 使用 JavaScript 渲染并嵌入为 Markdown 图片
+                chat_id = self._extract_chat_id(body, __metadata__)
+                message_id = self._extract_message_id(body, __metadata__)
+
+                await self._emit_status(
+                    __event_emitter__,
+                    "思维导图: 正在渲染图片...",
+                    False,
+                )
+
+                if __event_call__:
+                    js_code = self._generate_image_js_code(
+                        unique_id=unique_id,
+                        chat_id=chat_id,
+                        message_id=message_id,
+                        markdown_syntax=markdown_syntax,
+                    )
+
+                    await __event_call__(
+                        {
+                            "type": "execute",
+                            "data": {"code": js_code},
+                        }
+                    )
+
+                await self._emit_status(
+                    __event_emitter__, "思维导图: 图片已生成!", True
+                )
+                await self._emit_notification(
+                    __event_emitter__,
+                    f"思维导图图片已生成,{user_name}!",
+                    "success",
+                )
+                logger.info("Action: 思维导图 (v0.9.1) 图片模式完成")
+                return body
+
+            # HTML 模式(默认): 嵌入为 HTML 块
             html_embed_tag = f"```html\n{final_html}\n```"
             body["messages"][-1]["content"] = f"{long_text_content}\n\n{html_embed_tag}"
 
-            await self._emit_status(__event_emitter__, "思维导图: 绘制完成！", True)
+            await self._emit_status(__event_emitter__, "思维导图: 绘制完成!", True)
             await self._emit_notification(
-                __event_emitter__, f"思维导图已生成，{user_name}！", "success"
+                __event_emitter__, f"思维导图已生成,{user_name}!", "success"
             )
-            logger.info("Action: 思维导图 (v12) completed successfully")
+            logger.info("Action: 思维导图 (v0.9.1) HTML 模式完成")
 
         except Exception as e:
             error_message = f"思维导图处理失败: {str(e)}"
             logger.error(f"思维导图错误: {error_message}", exc_info=True)
-            user_facing_error = f"抱歉，思维导图在处理时遇到错误: {str(e)}。\n请检查Open WebUI后端日志获取更多详情。"
+            user_facing_error = f"抱歉,思维导图在处理时遇到错误: {str(e)}。\n请检查Open WebUI后端日志获取更多详情。"
             body["messages"][-1][
                 "content"
             ] = f"{long_text_content}\n\n❌ **错误:** {user_facing_error}"
 
             await self._emit_status(__event_emitter__, "思维导图: 处理失败。", True)
             await self._emit_notification(
-                __event_emitter__, f"思维导图生成失败, {user_name}！", "error"
+                __event_emitter__, f"思维导图生成失败, {user_name}!", "error"
             )
 
         return body

plugins/actions/summary/README.md

@@ -1,24 +0,0 @@
-# Deep Reading & Summary
-
-A powerful tool for analyzing long texts, generating detailed summaries, key points, and actionable insights.
-
-## Features
-
-- **Deep Analysis**: Goes beyond simple summarization to understand the core message.
-- **Key Point Extraction**: Identifies and lists the most important information.
-- **Actionable Advice**: Provides practical suggestions based on the text content.
-
-## Usage
-
-1. Install the plugin.
-2. Send a long text or article to the chat.
-3. Click the "Deep Reading" button (or trigger via command).
-
-## Author
-
-Fu-Jie
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## License
-
-MIT License

plugins/actions/summary/README_CN.md

@@ -1,24 +0,0 @@
-# 深度阅读与摘要 (Deep Reading & Summary)
-
-一个强大的长文本分析工具，用于生成详细摘要、关键信息点和可执行的行动建议。
-
-## 功能特点
-
-- **深度分析**：超越简单的总结，深入理解核心信息。
-- **关键点提取**：识别并列出最重要的信息点。
-- **行动建议**：基于文本内容提供切实可行的建议。
-
-## 使用方法
-
-1. 安装插件。
-2. 发送长文本或文章到聊天框。
-3. 点击“精读”按钮（或通过命令触发）。
-
-## 作者
-
-Fu-Jie
-GitHub: [Fu-Jie/awesome-openwebui](https://github.com/Fu-Jie/awesome-openwebui)
-
-## 许可证
-
-MIT License

plugins/actions/summary/summary.py

@@ -1,676 +0,0 @@
-"""
-title: Deep Reading & Summary
-author: Fu-Jie
-author_url: https://github.com/Fu-Jie
-funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.1.0
-icon_url: data:image/svg+xml;base64,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
-description: Provides deep reading analysis and summarization for long texts.
-requirements: jinja2, markdown
-"""
-
-from pydantic import BaseModel, Field
-from typing import Optional, Dict, Any
-import logging
-import re
-from fastapi import Request
-from datetime import datetime
-import pytz
-import markdown
-from jinja2 import Template
-
-from open_webui.utils.chat import generate_chat_completion
-from open_webui.models.users import Users
-
-logging.basicConfig(
-    level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
-)
-logger = logging.getLogger(__name__)
-
-# =================================================================
-# HTML Wrapper Template (supports multiple plugins and grid layout)
-# =================================================================
-HTML_WRAPPER_TEMPLATE = """
-<!-- OPENWEBUI_PLUGIN_OUTPUT -->
-<!DOCTYPE html>
-<html lang="{user_language}">
-<head>
-    <meta charset="UTF-8">
-    <meta name="viewport" content="width=device-width, initial-scale=1.0">
-    <style>
-        body { 
-            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; 
-            margin: 0; 
-            padding: 10px; 
-            background-color: transparent; 
-        }
-        #main-container { 
-            display: flex; 
-            flex-wrap: wrap; 
-            gap: 20px; 
-            align-items: flex-start; 
-            width: 100%;
-        }
-        .plugin-item { 
-            flex: 1 1 400px; /* Default width, allows shrinking/growing */
-            min-width: 300px; 
-            background: white; 
-            border-radius: 12px; 
-            box-shadow: 0 4px 6px rgba(0,0,0,0.05); 
-            overflow: hidden; 
-            border: 1px solid #e5e7eb; 
-            transition: all 0.3s ease;
-        }
-        .plugin-item:hover {
-            box-shadow: 0 10px 15px rgba(0,0,0,0.1);
-        }
-        @media (max-width: 768px) { 
-            .plugin-item { flex: 1 1 100%; } 
-        }
-        /* STYLES_INSERTION_POINT */
-    </style>
-</head>
-<body>
-    <div id="main-container">
-        <!-- CONTENT_INSERTION_POINT -->
-    </div>
-    <!-- SCRIPTS_INSERTION_POINT -->
-</body>
-</html>
-"""
-
-# =================================================================
-# Internal LLM Prompts
-# =================================================================
-
-SYSTEM_PROMPT_READING_ASSISTANT = """
-You are a professional Deep Text Analysis Expert, specializing in reading long texts and extracting the essence. Your task is to conduct a comprehensive and in-depth analysis.
-
-Please provide the following:
-1.  **Detailed Summary**: Summarize the core content of the text in 2-3 paragraphs, ensuring accuracy and completeness. Do not be too brief; ensure the reader fully understands the main idea.
-2.  **Key Information Points**: List 5-8 most important facts, viewpoints, or arguments. Each point should:
-    - Be specific and insightful
-    - Include necessary details and context
-    - Use Markdown list format
-3.  **Actionable Advice**: Identify and refine specific, actionable items from the text. Each suggestion should:
-    - Be clear and actionable
-    - Include execution priority or timing suggestions
-    - If there are no clear action items, provide learning suggestions or thinking directions
-
-Please strictly follow these guidelines:
--   **Language**: All output must be in the user's specified language.
--   **Format**: Please strictly follow the Markdown format below, ensuring each section has a clear header:
-    ## Summary
-    [Detailed summary content here, 2-3 paragraphs, use Markdown **bold** or *italic* to emphasize key points]
-
-    ## Key Information Points
-    - [Key Point 1: Include specific details and context]
-    - [Key Point 2: Include specific details and context]
-    - [Key Point 3: Include specific details and context]
-    - [At least 5, at most 8 key points]
-
-    ## Actionable Advice
-    - [Action Item 1: Specific, actionable, include priority]
-    - [Action Item 2: Specific, actionable, include priority]
-    - [If no clear action items, provide learning suggestions or thinking directions]
--   **Depth First**: Analysis should be deep and comprehensive, not superficial.
--   **Action Oriented**: Focus on actionable suggestions and next steps.
--   **Analysis Results Only**: Do not include any extra pleasantries, explanations, or leading text.
-"""
-
-USER_PROMPT_GENERATE_SUMMARY = """
-Please conduct a deep analysis of the following long text, providing:
-1.  Detailed Summary (2-3 paragraphs, comprehensive overview)
-2.  Key Information Points List (5-8 items, including specific details)
-3.  Actionable Advice (Specific, clear, including priority)
-
----
-**User Context:**
-User Name: {user_name}
-Current Date/Time: {current_date_time_str}
-Weekday: {current_weekday}
-Timezone: {current_timezone_str}
-User Language: {user_language}
----
-
-**Long Text Content:**
-```
-{long_text_content}
-```
-
-Please conduct a deep and comprehensive analysis, focusing on actionable advice.
-"""
-
-# =================================================================
-# Frontend HTML Template (Jinja2 Syntax)
-# =================================================================
-
-CSS_TEMPLATE_SUMMARY = """
-        :root {
-            --primary-color: #4285f4;
-            --secondary-color: #1e88e5;
-            --action-color: #34a853;
-            --background-color: #f8f9fa;
-            --card-bg-color: #ffffff;
-            --text-color: #202124;
-            --muted-text-color: #5f6368;
-            --border-color: #dadce0;
-            --header-gradient: linear-gradient(135deg, #4285f4, #1e88e5);
-            --shadow: 0 1px 3px rgba(60,64,67,.3);
-            --border-radius: 8px;
-            --font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "Helvetica Neue", Arial, sans-serif;
-        }
-        .summary-container-wrapper {
-            font-family: var(--font-family);
-            line-height: 1.8;
-            color: var(--text-color);
-            height: 100%;
-            display: flex;
-            flex-direction: column;
-        }
-        .summary-container-wrapper .header {
-            background: var(--header-gradient);
-            color: white;
-            padding: 20px 24px;
-            text-align: center;
-        }
-        .summary-container-wrapper .header h1 {
-            margin: 0;
-            font-size: 1.5em;
-            font-weight: 500;
-            letter-spacing: -0.5px;
-        }
-        .summary-container-wrapper .user-context {
-            font-size: 0.8em;
-            color: var(--muted-text-color);
-            background-color: #f1f3f4;
-            padding: 8px 16px;
-            display: flex;
-            justify-content: space-around;
-            flex-wrap: wrap;
-            border-bottom: 1px solid var(--border-color);
-        }
-        .summary-container-wrapper .user-context span { margin: 2px 8px; }
-        .summary-container-wrapper .content { padding: 20px; flex-grow: 1; }
-        .summary-container-wrapper .section {
-            margin-bottom: 16px;
-            padding-bottom: 16px;
-            border-bottom: 1px solid #e8eaed;
-        }
-        .summary-container-wrapper .section:last-child {
-            border-bottom: none;
-            margin-bottom: 0;
-            padding-bottom: 0;
-        }
-        .summary-container-wrapper .section h2 {
-            margin-top: 0;
-            margin-bottom: 12px;
-            font-size: 1.2em;
-            font-weight: 500;
-            color: var(--text-color);
-            display: flex;
-            align-items: center;
-            padding-bottom: 8px;
-            border-bottom: 2px solid var(--primary-color);
-        }
-        .summary-container-wrapper .section h2 .icon {
-            margin-right: 8px;
-            font-size: 1.1em;
-            line-height: 1;
-        }
-        .summary-container-wrapper .summary-section h2 { border-bottom-color: var(--primary-color); }
-        .summary-container-wrapper .keypoints-section h2 { border-bottom-color: var(--secondary-color); }
-        .summary-container-wrapper .actions-section h2 { border-bottom-color: var(--action-color); }
-        .summary-container-wrapper .html-content {
-            font-size: 0.95em;
-            line-height: 1.7;
-        }
-        .summary-container-wrapper .html-content p:first-child { margin-top: 0; }
-        .summary-container-wrapper .html-content p:last-child { margin-bottom: 0; }
-        .summary-container-wrapper .html-content ul {
-            list-style: none;
-            padding-left: 0;
-            margin: 12px 0;
-        }
-        .summary-container-wrapper .html-content li {
-            padding: 8px 0 8px 24px;
-            position: relative;
-            margin-bottom: 6px;
-            line-height: 1.6;
-        }
-        .summary-container-wrapper .html-content li::before {
-            position: absolute;
-            left: 0;
-            top: 8px;
-            font-family: 'Arial';
-            font-weight: bold;
-            font-size: 1em;
-        }
-        .summary-container-wrapper .keypoints-section .html-content li::before { 
-            content: '•'; 
-            color: var(--secondary-color);
-            font-size: 1.3em;
-            top: 5px;
-        }
-        .summary-container-wrapper .actions-section .html-content li::before { 
-            content: '▸'; 
-            color: var(--action-color); 
-        }
-        .summary-container-wrapper .no-content { 
-            color: var(--muted-text-color); 
-            font-style: italic;
-            padding: 12px;
-            background: #f8f9fa;
-            border-radius: 4px;
-        }
-        .summary-container-wrapper .footer {
-            text-align: center;
-            padding: 16px;
-            font-size: 0.8em;
-            color: #5f6368;
-            background-color: #f8f9fa;
-            border-top: 1px solid var(--border-color);
-        }
-"""
-
-CONTENT_TEMPLATE_SUMMARY = """
-        <div class="summary-container-wrapper">
-            <div class="header">
-                <h1>📖 Deep Reading: Analysis Report</h1>
-            </div>
-            <div class="user-context">
-                <span><strong>User:</strong> {user_name}</span>
-                <span><strong>Time:</strong> {current_date_time_str}</span>
-            </div>
-            <div class="content">
-                <div class="section summary-section">
-                    <h2><span class="icon">📝</span>Detailed Summary</h2>
-                    <div class="html-content">{summary_html}</div>
-                </div>
-                <div class="section keypoints-section">
-                    <h2><span class="icon">💡</span>Key Information Points</h2>
-                    <div class="html-content">{keypoints_html}</div>
-                </div>
-                <div class="section actions-section">
-                    <h2><span class="icon">🎯</span>Actionable Advice</h2>
-                    <div class="html-content">{actions_html}</div>
-                </div>
-            </div>
-            <div class="footer">
-                <p>&copy; {current_year} Deep Reading - Text Analysis Service</p>
-            </div>
-        </div>
-"""
-
-
-class Action:
-    class Valves(BaseModel):
-        SHOW_STATUS: bool = Field(
-            default=True,
-            description="Whether to show operation status updates in the chat interface.",
-        )
-        MODEL_ID: str = Field(
-            default="",
-            description="Built-in LLM Model ID used for text analysis. If empty, uses the current conversation's model.",
-        )
-        MIN_TEXT_LENGTH: int = Field(
-            default=200,
-            description="Minimum text length required for deep analysis (characters). Recommended 200+.",
-        )
-        RECOMMENDED_MIN_LENGTH: int = Field(
-            default=500,
-            description="Recommended minimum text length for best analysis results.",
-        )
-        CLEAR_PREVIOUS_HTML: bool = Field(
-            default=False,
-            description="Whether to force clear previous plugin results (if True, overwrites instead of merging).",
-        )
-        MESSAGE_COUNT: int = Field(
-            default=1,
-            description="Number of recent messages to use for generation. Set to 1 for just the last message, or higher for more context.",
-        )
-
-    def __init__(self):
-        self.valves = self.Valves()
-
-    def _process_llm_output(self, llm_output: str) -> Dict[str, str]:
-        """
-        Parse LLM Markdown output and convert to HTML fragments.
-        """
-        summary_match = re.search(
-            r"##\s*Summary\s*\n(.*?)(?=\n##|$)", llm_output, re.DOTALL | re.IGNORECASE
-        )
-        keypoints_match = re.search(
-            r"##\s*Key Information Points\s*\n(.*?)(?=\n##|$)",
-            llm_output,
-            re.DOTALL | re.IGNORECASE,
-        )
-        actions_match = re.search(
-            r"##\s*Actionable Advice\s*\n(.*?)(?=\n##|$)",
-            llm_output,
-            re.DOTALL | re.IGNORECASE,
-        )
-
-        summary_md = summary_match.group(1).strip() if summary_match else ""
-        keypoints_md = keypoints_match.group(1).strip() if keypoints_match else ""
-        actions_md = actions_match.group(1).strip() if actions_match else ""
-
-        if not any([summary_md, keypoints_md, actions_md]):
-            summary_md = llm_output.strip()
-            logger.warning(
-                "LLM output did not follow expected Markdown format. Treating entire output as summary."
-            )
-
-        # Use 'nl2br' extension to convert newlines \n to <br>
-        md_extensions = ["nl2br"]
-        summary_html = (
-            markdown.markdown(summary_md, extensions=md_extensions)
-            if summary_md
-            else '<p class="no-content">Failed to extract summary.</p>'
-        )
-        keypoints_html = (
-            markdown.markdown(keypoints_md, extensions=md_extensions)
-            if keypoints_md
-            else '<p class="no-content">Failed to extract key information points.</p>'
-        )
-        actions_html = (
-            markdown.markdown(actions_md, extensions=md_extensions)
-            if actions_md
-            else '<p class="no-content">No explicit actionable advice.</p>'
-        )
-
-        return {
-            "summary_html": summary_html,
-            "keypoints_html": keypoints_html,
-            "actions_html": actions_html,
-        }
-
-    async def _emit_status(self, emitter, description: str, done: bool = False):
-        """Emits a status update event."""
-        if self.valves.SHOW_STATUS and emitter:
-            await emitter(
-                {"type": "status", "data": {"description": description, "done": done}}
-            )
-
-    async def _emit_notification(self, emitter, content: str, ntype: str = "info"):
-        """Emits a notification event (info/success/warning/error)."""
-        if emitter:
-            await emitter(
-                {"type": "notification", "data": {"type": ntype, "content": content}}
-            )
-
-    def _remove_existing_html(self, content: str) -> str:
-        """Removes existing plugin-generated HTML code blocks from the content."""
-        pattern = r"```html\s*<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?```"
-        return re.sub(pattern, "", content).strip()
-
-    def _extract_text_content(self, content) -> str:
-        """Extract text from message content, supporting multimodal message formats"""
-        if isinstance(content, str):
-            return content
-        elif isinstance(content, list):
-            # Multimodal message: [{"type": "text", "text": "..."}, {"type": "image_url", ...}]
-            text_parts = []
-            for item in content:
-                if isinstance(item, dict) and item.get("type") == "text":
-                    text_parts.append(item.get("text", ""))
-                elif isinstance(item, str):
-                    text_parts.append(item)
-            return "\n".join(text_parts)
-        return str(content) if content else ""
-
-    def _merge_html(
-        self,
-        existing_html_code: str,
-        new_content: str,
-        new_styles: str = "",
-        new_scripts: str = "",
-        user_language: str = "en-US",
-    ) -> str:
-        """
-        Merges new content into an existing HTML container, or creates a new one.
-        """
-        if (
-            "<!-- OPENWEBUI_PLUGIN_OUTPUT -->" in existing_html_code
-            and "<!-- CONTENT_INSERTION_POINT -->" in existing_html_code
-        ):
-            base_html = existing_html_code
-            base_html = re.sub(r"^```html\s*", "", base_html)
-            base_html = re.sub(r"\s*```$", "", base_html)
-        else:
-            base_html = HTML_WRAPPER_TEMPLATE.replace("{user_language}", user_language)
-
-        wrapped_content = f'<div class="plugin-item">\n{new_content}\n</div>'
-
-        if new_styles:
-            base_html = base_html.replace(
-                "/* STYLES_INSERTION_POINT */",
-                f"{new_styles}\n/* STYLES_INSERTION_POINT */",
-            )
-
-        base_html = base_html.replace(
-            "<!-- CONTENT_INSERTION_POINT -->",
-            f"{wrapped_content}\n<!-- CONTENT_INSERTION_POINT -->",
-        )
-
-        if new_scripts:
-            base_html = base_html.replace(
-                "<!-- SCRIPTS_INSERTION_POINT -->",
-                f"{new_scripts}\n<!-- SCRIPTS_INSERTION_POINT -->",
-            )
-
-        return base_html.strip()
-
-    def _build_content_html(self, context: dict) -> str:
-        """
-        Build content HTML using context data.
-        """
-        return (
-            CONTENT_TEMPLATE_SUMMARY.replace(
-                "{user_name}", context.get("user_name", "User")
-            )
-            .replace(
-                "{current_date_time_str}", context.get("current_date_time_str", "")
-            )
-            .replace("{current_year}", context.get("current_year", ""))
-            .replace("{summary_html}", context.get("summary_html", ""))
-            .replace("{keypoints_html}", context.get("keypoints_html", ""))
-            .replace("{actions_html}", context.get("actions_html", ""))
-        )
-
-    async def action(
-        self,
-        body: dict,
-        __user__: Optional[Dict[str, Any]] = None,
-        __event_emitter__: Optional[Any] = None,
-        __request__: Optional[Request] = None,
-    ) -> Optional[dict]:
-        logger.info("Action: Deep Reading Started (v2.0.0)")
-
-        if isinstance(__user__, (list, tuple)):
-            user_language = (
-                __user__[0].get("language", "en-US") if __user__ else "en-US"
-            )
-            user_name = __user__[0].get("name", "User") if __user__[0] else "User"
-            user_id = (
-                __user__[0]["id"]
-                if __user__ and "id" in __user__[0]
-                else "unknown_user"
-            )
-        elif isinstance(__user__, dict):
-            user_language = __user__.get("language", "en-US")
-            user_name = __user__.get("name", "User")
-            user_id = __user__.get("id", "unknown_user")
-
-        now = datetime.now()
-        current_date_time_str = now.strftime("%B %d, %Y %H:%M:%S")
-        current_weekday = now.strftime("%A")
-        current_year = now.strftime("%Y")
-        current_timezone_str = "Unknown Timezone"
-
-        original_content = ""
-        try:
-            messages = body.get("messages", [])
-            if not messages:
-                raise ValueError("Unable to get valid user message content.")
-
-            # Get last N messages based on MESSAGE_COUNT
-            message_count = min(self.valves.MESSAGE_COUNT, len(messages))
-            recent_messages = messages[-message_count:]
-
-            # Aggregate content from selected messages with labels
-            aggregated_parts = []
-            for i, msg in enumerate(recent_messages, 1):
-                text_content = self._extract_text_content(msg.get("content"))
-                if text_content:
-                    role = msg.get("role", "unknown")
-                    role_label = (
-                        "User"
-                        if role == "user"
-                        else "Assistant" if role == "assistant" else role
-                    )
-                    aggregated_parts.append(
-                        f"[{role_label} Message {i}]\n{text_content}"
-                    )
-
-            if not aggregated_parts:
-                raise ValueError("Unable to get valid user message content.")
-
-            original_content = "\n\n---\n\n".join(aggregated_parts)
-
-            if len(original_content) < self.valves.MIN_TEXT_LENGTH:
-                short_text_message = f"Text content too short ({len(original_content)} chars), recommended at least {self.valves.MIN_TEXT_LENGTH} chars for effective deep analysis.\n\n💡 Tip: For short texts, consider using '⚡ Flash Card' for quick refinement."
-                await self._emit_notification(
-                    __event_emitter__, short_text_message, "warning"
-                )
-                return {
-                    "messages": [
-                        {"role": "assistant", "content": f"⚠️ {short_text_message}"}
-                    ]
-                }
-
-            # Recommend for longer texts
-            if len(original_content) < self.valves.RECOMMENDED_MIN_LENGTH:
-                await self._emit_notification(
-                    __event_emitter__,
-                    f"Text length is {len(original_content)} chars. Recommended {self.valves.RECOMMENDED_MIN_LENGTH}+ chars for best analysis results.",
-                    "info",
-                )
-
-            await self._emit_notification(
-                __event_emitter__,
-                "📖 Deep Reading started, analyzing deeply...",
-                "info",
-            )
-            await self._emit_status(
-                __event_emitter__,
-                "📖 Deep Reading: Analyzing text, extracting essence...",
-                False,
-            )
-
-            formatted_user_prompt = USER_PROMPT_GENERATE_SUMMARY.format(
-                user_name=user_name,
-                current_date_time_str=current_date_time_str,
-                current_weekday=current_weekday,
-                current_timezone_str=current_timezone_str,
-                user_language=user_language,
-                long_text_content=original_content,
-            )
-
-            # Determine model to use
-            target_model = self.valves.MODEL_ID
-            if not target_model:
-                target_model = body.get("model")
-
-            llm_payload = {
-                "model": target_model,
-                "messages": [
-                    {"role": "system", "content": SYSTEM_PROMPT_READING_ASSISTANT},
-                    {"role": "user", "content": formatted_user_prompt},
-                ],
-                "stream": False,
-            }
-
-            user_obj = Users.get_user_by_id(user_id)
-            if not user_obj:
-                raise ValueError(f"Unable to get user object, User ID: {user_id}")
-
-            llm_response = await generate_chat_completion(
-                __request__, llm_payload, user_obj
-            )
-            assistant_response_content = llm_response["choices"][0]["message"][
-                "content"
-            ]
-
-            processed_content = self._process_llm_output(assistant_response_content)
-
-            context = {
-                "user_language": user_language,
-                "user_name": user_name,
-                "current_date_time_str": current_date_time_str,
-                "current_weekday": current_weekday,
-                "current_year": current_year,
-                **processed_content,
-            }
-
-            content_html = self._build_content_html(context)
-
-            # Extract existing HTML if any
-            existing_html_block = ""
-            match = re.search(
-                r"```html\s*(<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?)```",
-                original_content,
-            )
-            if match:
-                existing_html_block = match.group(1)
-
-            if self.valves.CLEAR_PREVIOUS_HTML:
-                original_content = self._remove_existing_html(original_content)
-                final_html = self._merge_html(
-                    "", content_html, CSS_TEMPLATE_SUMMARY, "", user_language
-                )
-            else:
-                if existing_html_block:
-                    original_content = self._remove_existing_html(original_content)
-                    final_html = self._merge_html(
-                        existing_html_block,
-                        content_html,
-                        CSS_TEMPLATE_SUMMARY,
-                        "",
-                        user_language,
-                    )
-                else:
-                    final_html = self._merge_html(
-                        "", content_html, CSS_TEMPLATE_SUMMARY, "", user_language
-                    )
-
-            html_embed_tag = f"```html\n{final_html}\n```"
-            body["messages"][-1]["content"] = f"{original_content}\n\n{html_embed_tag}"
-
-            await self._emit_status(
-                __event_emitter__, "📖 Deep Reading: Analysis complete!", True
-            )
-            await self._emit_notification(
-                __event_emitter__,
-                f"📖 Deep Reading complete, {user_name}! Deep analysis report generated.",
-                "success",
-            )
-
-        except Exception as e:
-            error_message = f"Deep Reading processing failed: {str(e)}"
-            logger.error(f"Deep Reading Error: {error_message}", exc_info=True)
-            user_facing_error = f"Sorry, Deep Reading encountered an error while processing: {str(e)}.\nPlease check Open WebUI backend logs for more details."
-            body["messages"][-1][
-                "content"
-            ] = f"{original_content}\n\n❌ **Error:** {user_facing_error}"
-
-            await self._emit_status(
-                __event_emitter__, "Deep Reading: Processing failed.", True
-            )
-            await self._emit_notification(
-                __event_emitter__,
-                f"Deep Reading processing failed, {user_name}!",
-                "error",
-            )
-
-        return body

plugins/actions/summary/精读.py

@@ -1,663 +0,0 @@
-"""
-title: 精读 (Deep Reading)
-icon_url: data:image/svg+xml;base64,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
-version: 0.1.0
-description: 深度分析长篇文本，提炼详细摘要、关键信息点和可执行的行动建议，适合工作和学习场景。
-requirements: jinja2, markdown
-"""
-
-from pydantic import BaseModel, Field
-from typing import Optional, Dict, Any
-import logging
-import re
-from fastapi import Request
-from datetime import datetime
-import pytz
-import markdown
-from jinja2 import Template
-
-from open_webui.utils.chat import generate_chat_completion
-from open_webui.models.users import Users
-
-logging.basicConfig(
-    level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
-)
-logger = logging.getLogger(__name__)
-
-# =================================================================
-# HTML 容器模板 (支持多插件共存与网格布局)
-# =================================================================
-HTML_WRAPPER_TEMPLATE = """
-<!-- OPENWEBUI_PLUGIN_OUTPUT -->
-<!DOCTYPE html>
-<html lang="{user_language}">
-<head>
-    <meta charset="UTF-8">
-    <meta name="viewport" content="width=device-width, initial-scale=1.0">
-    <style>
-        body { 
-            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; 
-            margin: 0; 
-            padding: 10px; 
-            background-color: transparent; 
-        }
-        #main-container { 
-            display: flex; 
-            flex-wrap: wrap; 
-            gap: 20px; 
-            align-items: flex-start; 
-            width: 100%;
-        }
-        .plugin-item { 
-            flex: 1 1 400px; /* 默认宽度，允许伸缩 */
-            min-width: 300px; 
-            background: white; 
-            border-radius: 12px; 
-            box-shadow: 0 4px 6px rgba(0,0,0,0.05); 
-            overflow: hidden; 
-            border: 1px solid #e5e7eb; 
-            transition: all 0.3s ease;
-        }
-        .plugin-item:hover {
-            box-shadow: 0 10px 15px rgba(0,0,0,0.1);
-        }
-        @media (max-width: 768px) { 
-            .plugin-item { flex: 1 1 100%; } 
-        }
-        /* STYLES_INSERTION_POINT */
-    </style>
-</head>
-<body>
-    <div id="main-container">
-        <!-- CONTENT_INSERTION_POINT -->
-    </div>
-    <!-- SCRIPTS_INSERTION_POINT -->
-</body>
-</html>
-"""
-
-# =================================================================
-# 内部 LLM 提示词设计
-# =================================================================
-
-SYSTEM_PROMPT_READING_ASSISTANT = """
-你是一个专业的深度文本分析专家，擅长精读长篇文本并提炼精华。你的任务是进行全面、深入的分析。
-
-请提供以下内容：
-1.  **详细摘要**：用 2-3 段话全面总结文本的核心内容，确保准确性和完整性。不要过于简略，要让读者充分理解文本主旨。
-2.  **关键信息点**：列出 5-8 个最重要的事实、观点或论据。每个信息点应该：
-    - 具体且有深度
-    - 包含必要的细节和背景
-    - 使用 Markdown 列表格式
-3.  **行动建议**：从文本中识别并提炼出具体的、可执行的行动项。每个建议应该：
-    - 明确且可操作
-    - 包含执行的优先级或时间建议
-    - 如果没有明确的行动项，可以提供学习建议或思考方向
-
-请严格遵循以下指导原则：
--   **语言**：所有输出必须使用用户指定的语言。
--   **格式**：请严格按照以下 Markdown 格式输出，确保每个部分都有明确的标题：
-    ## 摘要
-    [这里是详细的摘要内容，2-3段话，可以使用 Markdown 进行**加粗**或*斜体*强调重点]
-
-    ## 关键信息点
-    - [关键点1：包含具体细节和背景]
-    - [关键点2：包含具体细节和背景]
-    - [关键点3：包含具体细节和背景]
-    - [至少5个，最多8个关键点]
-
-    ## 行动建议
-    - [行动项1：具体、可执行，包含优先级]
-    - [行动项2：具体、可执行，包含优先级]
-    - [如果没有明确行动项，提供学习建议或思考方向]
--   **深度优先**：分析要深入、全面，不要浮于表面。
--   **行动导向**：重点关注可执行的建议和下一步行动。
--   **只输出分析结果**：不要包含任何额外的寒暄、解释或引导性文字。
-"""
-
-USER_PROMPT_GENERATE_SUMMARY = """
-请对以下长篇文本进行深度分析，提供：
-1.  详细的摘要（2-3段话，全面概括文本内容）
-2.  关键信息点列表（5-8个，包含具体细节）
-3.  可执行的行动建议（具体、明确，包含优先级）
-
----
-**用户上下文信息:**
-用户姓名: {user_name}
-当前日期时间: {current_date_time_str}
-当前星期: {current_weekday}
-当前时区: {current_timezone_str}
-用户语言: {user_language}
----
-
-**长篇文本内容:**
-```
-{long_text_content}
-```
-
-请进行深入、全面的分析，重点关注可执行的行动建议。
-"""
-
-# =================================================================
-# 前端 HTML 模板 (Jinja2 语法)
-# =================================================================
-
-CSS_TEMPLATE_SUMMARY = """
-        :root {
-            --primary-color: #4285f4;
-            --secondary-color: #1e88e5;
-            --action-color: #34a853;
-            --background-color: #f8f9fa;
-            --card-bg-color: #ffffff;
-            --text-color: #202124;
-            --muted-text-color: #5f6368;
-            --border-color: #dadce0;
-            --header-gradient: linear-gradient(135deg, #4285f4, #1e88e5);
-            --shadow: 0 1px 3px rgba(60,64,67,.3);
-            --border-radius: 8px;
-            --font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "Helvetica Neue", Arial, sans-serif;
-        }
-        .summary-container-wrapper {
-            font-family: var(--font-family);
-            line-height: 1.8;
-            color: var(--text-color);
-            height: 100%;
-            display: flex;
-            flex-direction: column;
-        }
-        .summary-container-wrapper .header {
-            background: var(--header-gradient);
-            color: white;
-            padding: 20px 24px;
-            text-align: center;
-        }
-        .summary-container-wrapper .header h1 {
-            margin: 0;
-            font-size: 1.5em;
-            font-weight: 500;
-            letter-spacing: -0.5px;
-        }
-        .summary-container-wrapper .user-context {
-            font-size: 0.8em;
-            color: var(--muted-text-color);
-            background-color: #f1f3f4;
-            padding: 8px 16px;
-            display: flex;
-            justify-content: space-around;
-            flex-wrap: wrap;
-            border-bottom: 1px solid var(--border-color);
-        }
-        .summary-container-wrapper .user-context span { margin: 2px 8px; }
-        .summary-container-wrapper .content { padding: 20px; flex-grow: 1; }
-        .summary-container-wrapper .section {
-            margin-bottom: 16px;
-            padding-bottom: 16px;
-            border-bottom: 1px solid #e8eaed;
-        }
-        .summary-container-wrapper .section:last-child {
-            border-bottom: none;
-            margin-bottom: 0;
-            padding-bottom: 0;
-        }
-        .summary-container-wrapper .section h2 {
-            margin-top: 0;
-            margin-bottom: 12px;
-            font-size: 1.2em;
-            font-weight: 500;
-            color: var(--text-color);
-            display: flex;
-            align-items: center;
-            padding-bottom: 8px;
-            border-bottom: 2px solid var(--primary-color);
-        }
-        .summary-container-wrapper .section h2 .icon {
-            margin-right: 8px;
-            font-size: 1.1em;
-            line-height: 1;
-        }
-        .summary-container-wrapper .summary-section h2 { border-bottom-color: var(--primary-color); }
-        .summary-container-wrapper .keypoints-section h2 { border-bottom-color: var(--secondary-color); }
-        .summary-container-wrapper .actions-section h2 { border-bottom-color: var(--action-color); }
-        .summary-container-wrapper .html-content {
-            font-size: 0.95em;
-            line-height: 1.7;
-        }
-        .summary-container-wrapper .html-content p:first-child { margin-top: 0; }
-        .summary-container-wrapper .html-content p:last-child { margin-bottom: 0; }
-        .summary-container-wrapper .html-content ul {
-            list-style: none;
-            padding-left: 0;
-            margin: 12px 0;
-        }
-        .summary-container-wrapper .html-content li {
-            padding: 8px 0 8px 24px;
-            position: relative;
-            margin-bottom: 6px;
-            line-height: 1.6;
-        }
-        .summary-container-wrapper .html-content li::before {
-            position: absolute;
-            left: 0;
-            top: 8px;
-            font-family: 'Arial';
-            font-weight: bold;
-            font-size: 1em;
-        }
-        .summary-container-wrapper .keypoints-section .html-content li::before { 
-            content: '•'; 
-            color: var(--secondary-color);
-            font-size: 1.3em;
-            top: 5px;
-        }
-        .summary-container-wrapper .actions-section .html-content li::before { 
-            content: '▸'; 
-            color: var(--action-color); 
-        }
-        .summary-container-wrapper .no-content { 
-            color: var(--muted-text-color); 
-            font-style: italic;
-            padding: 12px;
-            background: #f8f9fa;
-            border-radius: 4px;
-        }
-        .summary-container-wrapper .footer {
-            text-align: center;
-            padding: 16px;
-            font-size: 0.8em;
-            color: #5f6368;
-            background-color: #f8f9fa;
-            border-top: 1px solid var(--border-color);
-        }
-"""
-
-CONTENT_TEMPLATE_SUMMARY = """
-        <div class="summary-container-wrapper">
-            <div class="header">
-                <h1>📖 精读：深度分析报告</h1>
-            </div>
-            <div class="user-context">
-                <span><strong>用户:</strong> {user_name}</span>
-                <span><strong>时间:</strong> {current_date_time_str}</span>
-            </div>
-            <div class="content">
-                <div class="section summary-section">
-                    <h2><span class="icon">📝</span>详细摘要</h2>
-                    <div class="html-content">{summary_html}</div>
-                </div>
-                <div class="section keypoints-section">
-                    <h2><span class="icon">💡</span>关键信息点</h2>
-                    <div class="html-content">{keypoints_html}</div>
-                </div>
-                <div class="section actions-section">
-                    <h2><span class="icon">🎯</span>行动建议</h2>
-                    <div class="html-content">{actions_html}</div>
-                </div>
-            </div>
-            <div class="footer">
-                <p>&copy; {current_year} 精读 - 深度文本分析服务</p>
-            </div>
-        </div>
-"""
-
-
-class Action:
-    class Valves(BaseModel):
-        SHOW_STATUS: bool = Field(
-            default=True, description="是否在聊天界面显示操作状态更新。"
-        )
-        MODEL_ID: str = Field(
-            default="",
-            description="用于文本分析的内置LLM模型ID。如果为空，则使用当前对话的模型。",
-        )
-        MIN_TEXT_LENGTH: int = Field(
-            default=200,
-            description="进行深度分析所需的最小文本长度(字符数)。建议200字符以上。",
-        )
-        RECOMMENDED_MIN_LENGTH: int = Field(
-            default=500, description="建议的最小文本长度，以获得最佳分析效果。"
-        )
-        CLEAR_PREVIOUS_HTML: bool = Field(
-            default=False,
-            description="是否强制清除旧的插件结果（如果为 True，则不合并，直接覆盖）。",
-        )
-        MESSAGE_COUNT: int = Field(
-            default=1,
-            description="用于生成的最近消息数量。设置为1仅使用最后一条消息，更大值可包含更多上下文。",
-        )
-
-    def __init__(self):
-        self.valves = self.Valves()
-        self.weekday_map = {
-            "Monday": "星期一",
-            "Tuesday": "星期二",
-            "Wednesday": "星期三",
-            "Thursday": "星期四",
-            "Friday": "星期五",
-            "Saturday": "星期六",
-            "Sunday": "星期日",
-        }
-
-    def _process_llm_output(self, llm_output: str) -> Dict[str, str]:
-        """
-        解析LLM的Markdown输出,将其转换为HTML片段。
-        """
-        summary_match = re.search(
-            r"##\s*摘要\s*\n(.*?)(?=\n##|$)", llm_output, re.DOTALL
-        )
-        keypoints_match = re.search(
-            r"##\s*关键信息点\s*\n(.*?)(?=\n##|$)", llm_output, re.DOTALL
-        )
-        actions_match = re.search(
-            r"##\s*行动建议\s*\n(.*?)(?=\n##|$)", llm_output, re.DOTALL
-        )
-
-        summary_md = summary_match.group(1).strip() if summary_match else ""
-        keypoints_md = keypoints_match.group(1).strip() if keypoints_match else ""
-        actions_md = actions_match.group(1).strip() if actions_match else ""
-
-        if not any([summary_md, keypoints_md, actions_md]):
-            summary_md = llm_output.strip()
-            logger.warning("LLM输出未遵循预期的Markdown格式。将整个输出视为摘要。")
-
-        # 使用 'nl2br' 扩展将换行符 \n 转换为 <br>
-        md_extensions = ["nl2br"]
-        summary_html = (
-            markdown.markdown(summary_md, extensions=md_extensions)
-            if summary_md
-            else '<p class="no-content">未能提取摘要信息。</p>'
-        )
-        keypoints_html = (
-            markdown.markdown(keypoints_md, extensions=md_extensions)
-            if keypoints_md
-            else '<p class="no-content">未能提取关键信息点。</p>'
-        )
-        actions_html = (
-            markdown.markdown(actions_md, extensions=md_extensions)
-            if actions_md
-            else '<p class="no-content">暂无明确的行动建议。</p>'
-        )
-
-        return {
-            "summary_html": summary_html,
-            "keypoints_html": keypoints_html,
-            "actions_html": actions_html,
-        }
-
-    async def _emit_status(self, emitter, description: str, done: bool = False):
-        """发送状态更新事件。"""
-        if self.valves.SHOW_STATUS and emitter:
-            await emitter(
-                {"type": "status", "data": {"description": description, "done": done}}
-            )
-
-    async def _emit_notification(self, emitter, content: str, ntype: str = "info"):
-        """发送通知事件 (info/success/warning/error)。"""
-        if emitter:
-            await emitter(
-                {"type": "notification", "data": {"type": ntype, "content": content}}
-            )
-
-    def _remove_existing_html(self, content: str) -> str:
-        """移除内容中已有的插件生成 HTML 代码块 (通过标记识别)。"""
-        pattern = r"```html\s*<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?```"
-        return re.sub(pattern, "", content).strip()
-
-    def _extract_text_content(self, content) -> str:
-        """从消息内容中提取文本，支持多模态消息格式"""
-        if isinstance(content, str):
-            return content
-        elif isinstance(content, list):
-            # 多模态消息: [{"type": "text", "text": "..."}, {"type": "image_url", ...}]
-            text_parts = []
-            for item in content:
-                if isinstance(item, dict) and item.get("type") == "text":
-                    text_parts.append(item.get("text", ""))
-                elif isinstance(item, str):
-                    text_parts.append(item)
-            return "\n".join(text_parts)
-        return str(content) if content else ""
-
-    def _merge_html(
-        self,
-        existing_html_code: str,
-        new_content: str,
-        new_styles: str = "",
-        new_scripts: str = "",
-        user_language: str = "zh-CN",
-    ) -> str:
-        """
-        将新内容合并到现有的 HTML 容器中，或者创建一个新的容器。
-        """
-        if (
-            "<!-- OPENWEBUI_PLUGIN_OUTPUT -->" in existing_html_code
-            and "<!-- CONTENT_INSERTION_POINT -->" in existing_html_code
-        ):
-            base_html = existing_html_code
-            base_html = re.sub(r"^```html\s*", "", base_html)
-            base_html = re.sub(r"\s*```$", "", base_html)
-        else:
-            base_html = HTML_WRAPPER_TEMPLATE.replace("{user_language}", user_language)
-
-        wrapped_content = f'<div class="plugin-item">\n{new_content}\n</div>'
-
-        if new_styles:
-            base_html = base_html.replace(
-                "/* STYLES_INSERTION_POINT */",
-                f"{new_styles}\n/* STYLES_INSERTION_POINT */",
-            )
-
-        base_html = base_html.replace(
-            "<!-- CONTENT_INSERTION_POINT -->",
-            f"{wrapped_content}\n<!-- CONTENT_INSERTION_POINT -->",
-        )
-
-        if new_scripts:
-            base_html = base_html.replace(
-                "<!-- SCRIPTS_INSERTION_POINT -->",
-                f"{new_scripts}\n<!-- SCRIPTS_INSERTION_POINT -->",
-            )
-
-        return base_html.strip()
-
-    def _build_content_html(self, context: dict) -> str:
-        """
-        使用上下文数据构建内容 HTML。
-        """
-        return (
-            CONTENT_TEMPLATE_SUMMARY.replace(
-                "{user_name}", context.get("user_name", "用户")
-            )
-            .replace(
-                "{current_date_time_str}", context.get("current_date_time_str", "")
-            )
-            .replace("{current_year}", context.get("current_year", ""))
-            .replace("{summary_html}", context.get("summary_html", ""))
-            .replace("{keypoints_html}", context.get("keypoints_html", ""))
-            .replace("{actions_html}", context.get("actions_html", ""))
-        )
-
-    async def action(
-        self,
-        body: dict,
-        __user__: Optional[Dict[str, Any]] = None,
-        __event_emitter__: Optional[Any] = None,
-        __request__: Optional[Request] = None,
-    ) -> Optional[dict]:
-        logger.info("Action: 精读启动 (v2.0.0 - Deep Reading)")
-
-        if isinstance(__user__, (list, tuple)):
-            user_language = (
-                __user__[0].get("language", "zh-CN") if __user__ else "zh-CN"
-            )
-            user_name = __user__[0].get("name", "用户") if __user__[0] else "用户"
-            user_id = (
-                __user__[0]["id"]
-                if __user__ and "id" in __user__[0]
-                else "unknown_user"
-            )
-        elif isinstance(__user__, dict):
-            user_language = __user__.get("language", "zh-CN")
-            user_name = __user__.get("name", "用户")
-            user_id = __user__.get("id", "unknown_user")
-
-        now = datetime.now()
-        current_date_time_str = now.strftime("%Y年%m月%d日 %H:%M:%S")
-        current_weekday_en = now.strftime("%A")
-        current_weekday = self.weekday_map.get(current_weekday_en, current_weekday_en)
-        current_year = now.strftime("%Y")
-        current_timezone_str = "未知时区"
-
-        original_content = ""
-        try:
-            messages = body.get("messages", [])
-            if not messages:
-                raise ValueError("无法获取有效的用户消息内容。")
-
-            # Get last N messages based on MESSAGE_COUNT
-            message_count = min(self.valves.MESSAGE_COUNT, len(messages))
-            recent_messages = messages[-message_count:]
-
-            # Aggregate content from selected messages with labels
-            aggregated_parts = []
-            for i, msg in enumerate(recent_messages, 1):
-                text_content = self._extract_text_content(msg.get("content"))
-                if text_content:
-                    role = msg.get("role", "unknown")
-                    role_label = (
-                        "用户"
-                        if role == "user"
-                        else "助手" if role == "assistant" else role
-                    )
-                    aggregated_parts.append(f"[{role_label} 消息 {i}]\n{text_content}")
-
-            if not aggregated_parts:
-                raise ValueError("无法获取有效的用户消息内容。")
-
-            original_content = "\n\n---\n\n".join(aggregated_parts)
-
-            if len(original_content) < self.valves.MIN_TEXT_LENGTH:
-                short_text_message = f"文本内容过短({len(original_content)}字符)，建议至少{self.valves.MIN_TEXT_LENGTH}字符以获得有效的深度分析。\n\n💡 提示：对于短文本，建议使用'⚡ 闪记卡'进行快速提炼。"
-                await self._emit_notification(
-                    __event_emitter__, short_text_message, "warning"
-                )
-                return {
-                    "messages": [
-                        {"role": "assistant", "content": f"⚠️ {short_text_message}"}
-                    ]
-                }
-
-            # Recommend for longer texts
-            if len(original_content) < self.valves.RECOMMENDED_MIN_LENGTH:
-                await self._emit_notification(
-                    __event_emitter__,
-                    f"文本长度为{len(original_content)}字符。建议{self.valves.RECOMMENDED_MIN_LENGTH}字符以上可获得更好的分析效果。",
-                    "info",
-                )
-
-            await self._emit_notification(
-                __event_emitter__, "📖 精读已启动，正在进行深度分析...", "info"
-            )
-            await self._emit_status(
-                __event_emitter__, "📖 精读: 深入分析文本，提炼精华...", False
-            )
-
-            formatted_user_prompt = USER_PROMPT_GENERATE_SUMMARY.format(
-                user_name=user_name,
-                current_date_time_str=current_date_time_str,
-                current_weekday=current_weekday,
-                current_timezone_str=current_timezone_str,
-                user_language=user_language,
-                long_text_content=original_content,
-            )
-
-            # 确定使用的模型
-            target_model = self.valves.MODEL_ID
-            if not target_model:
-                target_model = body.get("model")
-
-            llm_payload = {
-                "model": target_model,
-                "messages": [
-                    {"role": "system", "content": SYSTEM_PROMPT_READING_ASSISTANT},
-                    {"role": "user", "content": formatted_user_prompt},
-                ],
-                "stream": False,
-            }
-
-            user_obj = Users.get_user_by_id(user_id)
-            if not user_obj:
-                raise ValueError(f"无法获取用户对象, 用户ID: {user_id}")
-
-            llm_response = await generate_chat_completion(
-                __request__, llm_payload, user_obj
-            )
-            assistant_response_content = llm_response["choices"][0]["message"][
-                "content"
-            ]
-
-            processed_content = self._process_llm_output(assistant_response_content)
-
-            context = {
-                "user_language": user_language,
-                "user_name": user_name,
-                "current_date_time_str": current_date_time_str,
-                "current_weekday": current_weekday,
-                "current_year": current_year,
-                **processed_content,
-            }
-
-            content_html = self._build_content_html(context)
-
-            # Extract existing HTML if any
-            existing_html_block = ""
-            match = re.search(
-                r"```html\s*(<!-- OPENWEBUI_PLUGIN_OUTPUT -->[\s\S]*?)```",
-                original_content,
-            )
-            if match:
-                existing_html_block = match.group(1)
-
-            if self.valves.CLEAR_PREVIOUS_HTML:
-                original_content = self._remove_existing_html(original_content)
-                final_html = self._merge_html(
-                    "", content_html, CSS_TEMPLATE_SUMMARY, "", user_language
-                )
-            else:
-                if existing_html_block:
-                    original_content = self._remove_existing_html(original_content)
-                    final_html = self._merge_html(
-                        existing_html_block,
-                        content_html,
-                        CSS_TEMPLATE_SUMMARY,
-                        "",
-                        user_language,
-                    )
-                else:
-                    final_html = self._merge_html(
-                        "", content_html, CSS_TEMPLATE_SUMMARY, "", user_language
-                    )
-
-            html_embed_tag = f"```html\n{final_html}\n```"
-            body["messages"][-1]["content"] = f"{original_content}\n\n{html_embed_tag}"
-
-            await self._emit_status(__event_emitter__, "📖 精读: 分析完成!", True)
-            await self._emit_notification(
-                __event_emitter__,
-                f"📖 精读完成，{user_name}！深度分析报告已生成。",
-                "success",
-            )
-
-        except Exception as e:
-            error_message = f"精读处理失败: {str(e)}"
-            logger.error(f"精读错误: {error_message}", exc_info=True)
-            user_facing_error = f"抱歉, 精读在处理时遇到错误: {str(e)}。\n请检查Open WebUI后端日志获取更多详情。"
-            body["messages"][-1][
-                "content"
-            ] = f"{original_content}\n\n❌ **错误:** {user_facing_error}"
-
-            await self._emit_status(__event_emitter__, "精读: 处理失败。", True)
-            await self._emit_notification(
-                __event_emitter__, f"精读处理失败, {user_name}!", "error"
-            )
-
-        return body

plugins/filters/async-context-compression/README.md

@@ -1,15 +1,26 @@
 # Async Context Compression Filter
 
-**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.1.0 | **License:** MIT
+**Author:** [Fu-Jie](https://github.com/Fu-Jie) | **Version:** 1.1.3 | **License:** MIT
 
 This filter reduces token consumption in long conversations through intelligent summarization and message compression while keeping conversations coherent.
 
-## What's new in 1.1.0 
+## What's new in 1.1.3
+- **Improved Compatibility**: Changed summary injection role from `user` to `assistant` for better compatibility across different LLMs.
+- **Enhanced Stability**: Fixed a race condition in state management that could cause "inlet state not found" warnings in high-concurrency scenarios.
+- **Bug Fixes**: Corrected default model handling to prevent misleading logs when no model is specified.
+
+## What's new in 1.1.2
+
+- **Open WebUI v0.7.x Compatibility**: Resolved a critical database session binding error affecting Open WebUI v0.7.x users. The plugin now dynamically discovers the database engine and session context, ensuring compatibility across versions.
+- **Enhanced Error Reporting**: Errors during background summary generation are now reported via both the status bar and browser console.
+- **Robust Model Handling**: Improved handling of missing or invalid model IDs to prevent crashes.
+
+## What's new in 1.1.1
+
+- **Frontend Debugging**: Added `show_debug_log` option to print debug info to the browser console (F12).
+- **Optimized Compression**: Improved token calculation logic to prevent aggressive truncation of history, ensuring more context is retained.
+
 
-- Reuses Open WebUI's shared database connection by default (no custom engine or env vars required).
-- Token-based thresholds (`compression_threshold_tokens`, `max_context_tokens`) for safer long-context handling.
-- Per-model overrides via `model_thresholds` for mixed-model workflows.
-- Documentation now mirrors the latest async workflow and retention-first injection.
 
 ---
 
@@ -54,6 +65,7 @@ It is recommended to keep this filter early in the chain so it runs before filte
 | `summary_temperature`          | `0.3`    | Randomness for summary generation. Lower is more deterministic.                                                                                                       |
 | `model_thresholds`             | `{}`     | Per-model overrides for `compression_threshold_tokens` and `max_context_tokens` (useful for mixed models).                                                            |
 | `debug_mode`                   | `true`   | Log verbose debug info. Set to `false` in production.                                                                                                                 |
+| `show_debug_log`               | `false`  | Print debug logs to browser console (F12). Useful for frontend debugging.                                                                                             |
 
 ---
 

plugins/filters/async-context-compression/README_CN.md

@@ -1,17 +1,28 @@
 # 异步上下文压缩过滤器
 
-**作者:** [Fu-Jie](https://github.com/Fu-Jie) | **版本:** 1.2.0 | **许可证:** MIT
+**作者:** [Fu-Jie](https://github.com/Fu-Jie) | **版本:** 1.1.3 | **许可证:** MIT
 
 > **重要提示**：为了确保所有过滤器的可维护性和易用性，每个过滤器都应附带清晰、完整的文档，以确保其功能、配置和使用方法得到充分说明。
 
 本过滤器通过智能摘要和消息压缩技术，在保持对话连贯性的同时，显著降低长对话的 Token 消耗。
 
-## 1.1.0 版本更新
+## 1.1.3 版本更新
+- **兼容性提升**: 将摘要注入角色从 `user` 改为 `assistant`，以提高在不同 LLM 之间的兼容性。
+- **稳定性增强**: 修复了状态管理中的竞态条件，解决了高并发场景下可能出现的“无法获取 inlet 状态”警告。
+- **Bug 修复**: 修正了默认模型处理逻辑，防止在未指定模型时产生误导性日志。
+
+## 1.1.2 版本更新
+
+- **Open WebUI v0.7.x 兼容性**: 修复了影响 Open WebUI v0.7.x 用户的严重数据库会话绑定错误。插件现在动态发现数据库引擎和会话上下文，确保跨版本兼容性。
+- **增强错误报告**: 后台摘要生成过程中的错误现在会通过状态栏和浏览器控制台同时报告。
+- **健壮的模型处理**: 改进了对缺失或无效模型 ID 的处理，防止程序崩溃。
+
+## 1.1.1 版本更新
+
+- **前端调试**: 新增 `show_debug_log` 选项，支持在浏览器控制台 (F12) 打印调试信息。
+- **压缩优化**: 优化 Token 计算逻辑，防止历史记录被过度截断，保留更多上下文。
+
 
-- 默认复用 OpenWebUI 内置数据库连接，无需自建引擎、无需配置 `DATABASE_URL`。
-- 基于 Token 的阈值控制（`compression_threshold_tokens`、`max_context_tokens`），长上下文更安全。
-- 支持 `model_thresholds` 为不同模型设置专属阈值，适合混用多模型场景。
-- 文档同步最新异步工作流与“先保留再注入”策略。
 
 ---
 
@@ -94,6 +105,11 @@
 - **默认值**: `true`
 - **描述**: 是否在 Open WebUI 的控制台日志中打印详细的调试信息（如 Token 计数、压缩进度、数据库操作等）。生产环境建议设为 `false`。
 
+#### `show_debug_log`
+
+- **默认值**: `false`
+- **描述**: 是否在浏览器控制台 (F12) 打印调试日志。便于前端调试。
+
 ---
 
 ## 故障排除

plugins/filters/async-context-compression/async_context_compression.py

@@ -5,7 +5,8 @@ author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
 description: Reduces token consumption in long conversations while maintaining coherence through intelligent summarization and message compression.
-version: 1.1.0
+version: 1.1.3
+openwebui_id: b1655bc8-6de9-4cad-8cb5-a6f7829a02ce
 license: MIT
 
 ═══════════════════════════════════════════════════════════════════════════════
@@ -138,6 +139,10 @@ debug_mode
   Default: true
   Description: Prints detailed debug information to the log. Recommended to set to `false` in production.
 
+show_debug_log
+  Default: false
+  Description: Print debug logs to browser console (F12). Useful for frontend debugging.
+
 🔧 Deployment
 ═══════════════════════════════════════════════════════
 
@@ -244,6 +249,7 @@ import asyncio
 import json
 import hashlib
 import time
+import contextlib
 
 # Open WebUI built-in imports
 from open_webui.utils.chat import generate_chat_completion
@@ -252,9 +258,10 @@ from fastapi.requests import Request
 from open_webui.main import app as webui_app
 
 # Open WebUI internal database (re-use shared connection)
-from open_webui.internal.db import engine as owui_engine
-from open_webui.internal.db import Session as owui_Session
-from open_webui.internal.db import Base as owui_Base
+try:
+    from open_webui.internal import db as owui_db
+except ModuleNotFoundError:  # pragma: no cover - filter runs inside Open WebUI
+    owui_db = None
 
 # Try to import tiktoken
 try:
@@ -264,14 +271,91 @@ except ImportError:
 
 # Database imports
 from sqlalchemy import Column, String, Text, DateTime, Integer, inspect
+from sqlalchemy.orm import declarative_base, sessionmaker
+from sqlalchemy.engine import Engine
 from datetime import datetime
 
 
+def _discover_owui_engine(db_module: Any) -> Optional[Engine]:
+    """Discover the Open WebUI SQLAlchemy engine via provided db module helpers."""
+    if db_module is None:
+        return None
+
+    db_context = getattr(db_module, "get_db_context", None) or getattr(
+        db_module, "get_db", None
+    )
+    if callable(db_context):
+        try:
+            with db_context() as session:
+                try:
+                    return session.get_bind()
+                except AttributeError:
+                    return getattr(session, "bind", None) or getattr(
+                        session, "engine", None
+                    )
+        except Exception as exc:
+            print(f"[DB Discover] get_db_context failed: {exc}")
+
+    for attr in ("engine", "ENGINE", "bind", "BIND"):
+        candidate = getattr(db_module, attr, None)
+        if candidate is not None:
+            return candidate
+
+    return None
+
+
+def _discover_owui_schema(db_module: Any) -> Optional[str]:
+    """Discover the Open WebUI database schema name if configured."""
+    if db_module is None:
+        return None
+
+    try:
+        base = getattr(db_module, "Base", None)
+        metadata = getattr(base, "metadata", None) if base is not None else None
+        candidate = getattr(metadata, "schema", None) if metadata is not None else None
+        if isinstance(candidate, str) and candidate.strip():
+            return candidate.strip()
+    except Exception as exc:
+        print(f"[DB Discover] Base metadata schema lookup failed: {exc}")
+
+    try:
+        metadata_obj = getattr(db_module, "metadata_obj", None)
+        candidate = (
+            getattr(metadata_obj, "schema", None) if metadata_obj is not None else None
+        )
+        if isinstance(candidate, str) and candidate.strip():
+            return candidate.strip()
+    except Exception as exc:
+        print(f"[DB Discover] metadata_obj schema lookup failed: {exc}")
+
+    try:
+        from open_webui import env as owui_env
+
+        candidate = getattr(owui_env, "DATABASE_SCHEMA", None)
+        if isinstance(candidate, str) and candidate.strip():
+            return candidate.strip()
+    except Exception as exc:
+        print(f"[DB Discover] env schema lookup failed: {exc}")
+
+    return None
+
+
+owui_engine = _discover_owui_engine(owui_db)
+owui_schema = _discover_owui_schema(owui_db)
+owui_Base = getattr(owui_db, "Base", None) if owui_db is not None else None
+if owui_Base is None:
+    owui_Base = declarative_base()
+
+
 class ChatSummary(owui_Base):
     """Chat Summary Storage Table"""
 
     __tablename__ = "chat_summary"
-    __table_args__ = {"extend_existing": True}
+    __table_args__ = (
+        {"extend_existing": True, "schema": owui_schema}
+        if owui_schema
+        else {"extend_existing": True}
+    )
 
     id = Column(Integer, primary_key=True, autoincrement=True)
     chat_id = Column(String(255), unique=True, nullable=False, index=True)
@@ -284,14 +368,69 @@ class ChatSummary(owui_Base):
 class Filter:
     def __init__(self):
         self.valves = self.Valves()
+        self._owui_db = owui_db
         self._db_engine = owui_engine
-        self._SessionLocal = owui_Session
-        self.temp_state = {}  # Used to pass temporary data between inlet and outlet
+        self._db_engine = owui_engine
+        self._fallback_session_factory = (
+            sessionmaker(bind=self._db_engine) if self._db_engine else None
+        )
+        self._fallback_session_factory = (
+            sessionmaker(bind=self._db_engine) if self._db_engine else None
+        )
         self._init_database()
 
+    @contextlib.contextmanager
+    def _db_session(self):
+        """Yield a database session using Open WebUI helpers with graceful fallbacks."""
+        db_module = self._owui_db
+        db_context = None
+        if db_module is not None:
+            db_context = getattr(db_module, "get_db_context", None) or getattr(
+                db_module, "get_db", None
+            )
+
+        if callable(db_context):
+            with db_context() as session:
+                yield session
+                return
+
+        factory = None
+        if db_module is not None:
+            factory = getattr(db_module, "SessionLocal", None) or getattr(
+                db_module, "ScopedSession", None
+            )
+        if callable(factory):
+            session = factory()
+            try:
+                yield session
+            finally:
+                close = getattr(session, "close", None)
+                if callable(close):
+                    close()
+            return
+
+        if self._fallback_session_factory is None:
+            raise RuntimeError(
+                "Open WebUI database session is unavailable. Ensure Open WebUI's database layer is initialized."
+            )
+
+        session = self._fallback_session_factory()
+        try:
+            yield session
+        finally:
+            try:
+                session.close()
+            except Exception as exc:  # pragma: no cover - best-effort cleanup
+                print(f"[Database] ⚠️ Failed to close fallback session: {exc}")
+
     def _init_database(self):
         """Initializes the database table using Open WebUI's shared connection."""
         try:
+            if self._db_engine is None:
+                raise RuntimeError(
+                    "Open WebUI database engine is unavailable. Ensure Open WebUI is configured with a valid DATABASE_URL."
+                )
+
             # Check if table exists using SQLAlchemy inspect
             inspector = inspect(self._db_engine)
             if not inspector.has_table("chat_summary"):
@@ -354,11 +493,14 @@ class Filter:
         debug_mode: bool = Field(
             default=True, description="Enable detailed logging for debugging."
         )
+        show_debug_log: bool = Field(
+            default=False, description="Print debug logs to browser console (F12)"
+        )
 
     def _save_summary(self, chat_id: str, summary: str, compressed_count: int):
         """Saves the summary to the database."""
         try:
-            with self._SessionLocal() as session:
+            with self._db_session() as session:
                 # Find existing record
                 existing = session.query(ChatSummary).filter_by(chat_id=chat_id).first()
 
@@ -398,7 +540,7 @@ class Filter:
     def _load_summary_record(self, chat_id: str) -> Optional[ChatSummary]:
         """Loads the summary record object from the database."""
         try:
-            with self._SessionLocal() as session:
+            with self._db_session() as session:
                 record = session.query(ChatSummary).filter_by(chat_id=chat_id).first()
                 if record:
                     # Detach the object from the session so it can be used after session close
@@ -479,41 +621,121 @@ class Filter:
             "max_context_tokens": self.valves.max_context_tokens,
         }
 
-    def _inject_summary_to_first_message(self, message: dict, summary: str) -> dict:
-        """Injects the summary into the first message (prepended to content)."""
-        content = message.get("content", "")
-        summary_block = f"【Historical Conversation Summary】\n{summary}\n\n---\nBelow is the recent conversation:\n\n"
+    def _extract_chat_id(self, body: dict, metadata: Optional[dict]) -> str:
+        """Extract chat_id from body or metadata."""
+        if isinstance(body, dict):
+            chat_id = body.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
 
-        # Handle different content types
-        if isinstance(content, list):  # Multimodal content
-            # Find the first text part and insert the summary before it
-            new_content = []
-            summary_inserted = False
+            body_metadata = body.get("metadata", {})
+            if isinstance(body_metadata, dict):
+                chat_id = body_metadata.get("chat_id")
+                if isinstance(chat_id, str) and chat_id.strip():
+                    return chat_id.strip()
 
-            for part in content:
-                if (
-                    isinstance(part, dict)
-                    and part.get("type") == "text"
-                    and not summary_inserted
+        if isinstance(metadata, dict):
+            chat_id = metadata.get("chat_id")
+            if isinstance(chat_id, str) and chat_id.strip():
+                return chat_id.strip()
+
+        return ""
+
+    async def _emit_debug_log(
+        self,
+        __event_call__,
+        chat_id: str,
+        original_count: int,
+        compressed_count: int,
+        summary_length: int,
+        kept_first: int,
+        kept_last: int,
     ):
-                    # Prepend summary to the first text part
-                    new_content.append(
-                        {"type": "text", "text": summary_block + part.get("text", "")}
+        """Emit debug log to browser console via JS execution"""
+        if not self.valves.show_debug_log or not __event_call__:
+            return
+
+        try:
+            # Prepare data for JS
+            log_data = {
+                "chatId": chat_id,
+                "originalCount": original_count,
+                "compressedCount": compressed_count,
+                "summaryLength": summary_length,
+                "keptFirst": kept_first,
+                "keptLast": kept_last,
+                "ratio": (
+                    f"{(1 - compressed_count/original_count)*100:.1f}%"
+                    if original_count > 0
+                    else "0%"
+                ),
+            }
+
+            # Construct JS code
+            js_code = f"""
+                (async function() {{
+                    console.group("🗜️ Async Context Compression Debug");
+                    console.log("Chat ID:", {json.dumps(chat_id)});
+                    console.log("Messages:", {original_count} + " -> " + {compressed_count});
+                    console.log("Compression Ratio:", {json.dumps(log_data['ratio'])});
+                    console.log("Summary Length:", {summary_length} + " chars");
+                    console.log("Configuration:", {{
+                        "Keep First": {kept_first},
+                        "Keep Last": {kept_last}
+                    }});
+                    console.groupEnd();
+                }})();
+            """
+
+            await __event_call__(
+                {
+                    "type": "execute",
+                    "data": {"code": js_code},
+                }
             )
-                    summary_inserted = True
-                else:
-                    new_content.append(part)
+        except Exception as e:
+            print(f"Error emitting debug log: {e}")
 
-            # If no text part, insert at the beginning
-            if not summary_inserted:
-                new_content.insert(0, {"type": "text", "text": summary_block})
+    async def _log(self, message: str, type: str = "info", event_call=None):
+        """Unified logging to both backend (print) and frontend (console.log)"""
+        # Backend logging
+        if self.valves.debug_mode:
+            print(message)
 
-            message["content"] = new_content
+        # Frontend logging
+        if self.valves.show_debug_log and event_call:
+            try:
+                css = "color: #3b82f6;"  # Blue default
+                if type == "error":
+                    css = "color: #ef4444; font-weight: bold;"  # Red
+                elif type == "warning":
+                    css = "color: #f59e0b;"  # Orange
+                elif type == "success":
+                    css = "color: #10b981; font-weight: bold;"  # Green
 
-        elif isinstance(content, str):  # Plain text
-            message["content"] = summary_block + content
+                # Clean message for frontend: remove separators and extra newlines
+                lines = message.split("\n")
+                # Keep lines that don't start with lots of equals or hyphens
+                filtered_lines = [
+                    line
+                    for line in lines
+                    if not line.strip().startswith("====")
+                    and not line.strip().startswith("----")
+                ]
+                clean_message = "\n".join(filtered_lines).strip()
 
-        return message
+                if not clean_message:
+                    return
+
+                # Escape quotes in message for JS string
+                safe_message = clean_message.replace('"', '\\"').replace("\n", "\\n")
+
+                js_code = f"""
+                    console.log("%c[Compression] {safe_message}", "{css}");
+                """
+                await event_call({"type": "execute", "data": {"code": js_code}})
+            except Exception as e:
+                print(f"Failed to emit log to frontend: {e}")
 
     async def inlet(
         self,
@@ -521,35 +743,40 @@ class Filter:
         __user__: Optional[dict] = None,
         __metadata__: dict = None,
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> dict:
         """
         Executed before sending to the LLM.
         Compression Strategy: Only responsible for injecting existing summaries, no Token calculation.
         """
         messages = body.get("messages", [])
-        chat_id = __metadata__["chat_id"]
+        chat_id = self._extract_chat_id(body, __metadata__)
 
-        if self.valves.debug_mode:
-            print(f"\n{'='*60}")
-            print(f"[Inlet] Chat ID: {chat_id}")
-            print(f"[Inlet] Received {len(messages)} messages")
+        if not chat_id:
+            await self._log(
+                "[Inlet] ❌ Missing chat_id in metadata, skipping compression",
+                type="error",
+                event_call=__event_call__,
+            )
+            return body
+
+        if self.valves.debug_mode or self.valves.show_debug_log:
+            await self._log(
+                f"\n{'='*60}\n[Inlet] Chat ID: {chat_id}\n[Inlet] Received {len(messages)} messages",
+                event_call=__event_call__,
+            )
 
         # Record the target compression progress for the original messages, for use in outlet
         # Target is to compress up to the (total - keep_last) message
         target_compressed_count = max(0, len(messages) - self.valves.keep_last)
 
-        # [Optimization] Simple state cleanup check
-        if chat_id in self.temp_state:
-            if self.valves.debug_mode:
-                print(
-                    f"[Inlet] ⚠️ Overwriting unconsumed old state (Chat ID: {chat_id})"
-                )
+        # Record the target compression progress for the original messages, for use in outlet
+        # Target is to compress up to the (total - keep_last) message
+        target_compressed_count = max(0, len(messages) - self.valves.keep_last)
 
-        self.temp_state[chat_id] = target_compressed_count
-
-        if self.valves.debug_mode:
-            print(
-                f"[Inlet] Recorded target compression progress: {target_compressed_count}"
+        await self._log(
+            f"[Inlet] Recorded target compression progress: {target_compressed_count}",
+            event_call=__event_call__,
         )
 
         # Load summary record
@@ -578,7 +805,7 @@ class Filter:
                 f"---\n"
                 f"Below is the recent conversation:"
             )
-            summary_msg = {"role": "user", "content": summary_content}
+            summary_msg = {"role": "assistant", "content": summary_content}
 
             # 3. Tail messages (Tail) - All messages starting from the last compression point
             # Note: Must ensure head messages are not duplicated
@@ -599,9 +826,21 @@ class Filter:
                     }
                 )
 
-            if self.valves.debug_mode:
-                print(
-                    f"[Inlet] Applied summary: Head({len(head_messages)}) + Summary + Tail({len(tail_messages)})"
+            await self._log(
+                f"[Inlet] Applied summary: Head({len(head_messages)}) + Summary + Tail({len(tail_messages)})",
+                type="success",
+                event_call=__event_call__,
+            )
+
+            # Emit debug log to frontend (Keep the structured log as well)
+            await self._emit_debug_log(
+                __event_call__,
+                chat_id,
+                len(messages),
+                len(final_messages),
+                len(summary_record.summary),
+                self.valves.keep_first,
+                self.valves.keep_last,
             )
         else:
             # No summary, use original messages
@@ -609,9 +848,10 @@ class Filter:
 
         body["messages"] = final_messages
 
-        if self.valves.debug_mode:
-            print(f"[Inlet] Final send: {len(body['messages'])} messages")
-            print(f"{'='*60}\n")
+        await self._log(
+            f"[Inlet] Final send: {len(body['messages'])} messages\n{'='*60}\n",
+            event_call=__event_call__,
+        )
 
         return body
 
@@ -621,29 +861,50 @@ class Filter:
         __user__: Optional[dict] = None,
         __metadata__: dict = None,
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> dict:
         """
         Executed after the LLM response is complete.
         Calculates Token count in the background and triggers summary generation (does not block current response, does not affect content output).
         """
-        chat_id = __metadata__["chat_id"]
-        model = body.get("model", "gpt-3.5-turbo")
+        chat_id = self._extract_chat_id(body, __metadata__)
+        if not chat_id:
+            await self._log(
+                "[Outlet] ❌ Missing chat_id in metadata, skipping compression",
+                type="error",
+                event_call=__event_call__,
+            )
+            return body
+        model = body.get("model") or ""
 
-        if self.valves.debug_mode:
-            print(f"\n{'='*60}")
-            print(f"[Outlet] Chat ID: {chat_id}")
-            print(f"[Outlet] Response complete")
+        # Calculate target compression progress directly
+        # Assuming body['messages'] in outlet contains the full history (including new response)
+        messages = body.get("messages", [])
+        target_compressed_count = max(0, len(messages) - self.valves.keep_last)
+
+        if self.valves.debug_mode or self.valves.show_debug_log:
+            await self._log(
+                f"\n{'='*60}\n[Outlet] Chat ID: {chat_id}\n[Outlet] Response complete\n[Outlet] Calculated target compression progress: {target_compressed_count} (Messages: {len(messages)})",
+                event_call=__event_call__,
+            )
 
         # Process Token calculation and summary generation asynchronously in the background (do not wait for completion, do not affect output)
         asyncio.create_task(
             self._check_and_generate_summary_async(
-                chat_id, model, body, __user__, __event_emitter__
+                chat_id,
+                model,
+                body,
+                __user__,
+                target_compressed_count,
+                __event_emitter__,
+                __event_call__,
             )
         )
 
-        if self.valves.debug_mode:
-            print(f"[Outlet] Background processing started")
-            print(f"{'='*60}\n")
+        await self._log(
+            f"[Outlet] Background processing started\n{'='*60}\n",
+            event_call=__event_call__,
+        )
 
         return body
 
@@ -653,7 +914,9 @@ class Filter:
         model: str,
         body: dict,
         user_data: Optional[dict],
+        target_compressed_count: Optional[int],
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ):
         """
         Background processing: Calculates Token count and generates summary (does not block response).
@@ -667,36 +930,58 @@ class Filter:
                 "compression_threshold_tokens", self.valves.compression_threshold_tokens
             )
 
-            if self.valves.debug_mode:
-                print(f"\n[🔍 Background Calculation] Starting Token count...")
+            await self._log(
+                f"\n[🔍 Background Calculation] Starting Token count...",
+                event_call=__event_call__,
+            )
 
             # Calculate Token count in a background thread
             current_tokens = await asyncio.to_thread(
                 self._calculate_messages_tokens, messages
             )
 
-            if self.valves.debug_mode:
-                print(f"[🔍 Background Calculation] Token count: {current_tokens}")
+            await self._log(
+                f"[🔍 Background Calculation] Token count: {current_tokens}",
+                event_call=__event_call__,
+            )
 
             # Check if compression is needed
             if current_tokens >= compression_threshold_tokens:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🔍 Background Calculation] ⚡ Compression threshold triggered (Token: {current_tokens} >= {compression_threshold_tokens})"
+                await self._log(
+                    f"[🔍 Background Calculation] ⚡ Compression threshold triggered (Token: {current_tokens} >= {compression_threshold_tokens})",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
                 # Proceed to generate summary
                 await self._generate_summary_async(
-                    messages, chat_id, body, user_data, __event_emitter__
+                    messages,
+                    chat_id,
+                    body,
+                    user_data,
+                    target_compressed_count,
+                    __event_emitter__,
+                    __event_call__,
                 )
             else:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🔍 Background Calculation] Compression threshold not reached (Token: {current_tokens} < {compression_threshold_tokens})"
+                await self._log(
+                    f"[🔍 Background Calculation] Compression threshold not reached (Token: {current_tokens} < {compression_threshold_tokens})",
+                    event_call=__event_call__,
                 )
 
         except Exception as e:
-            print(f"[🔍 Background Calculation] ❌ Error: {str(e)}")
+            await self._log(
+                f"[🔍 Background Calculation] ❌ Error: {str(e)}",
+                type="error",
+                event_call=__event_call__,
+            )
+
+    def _clean_model_id(self, model_id: Optional[str]) -> Optional[str]:
+        """Cleans the model ID by removing whitespace and quotes."""
+        if not model_id:
+            return None
+        cleaned = model_id.strip().strip('"').strip("'")
+        return cleaned if cleaned else None
 
     async def _generate_summary_async(
         self,
@@ -704,7 +989,9 @@ class Filter:
         chat_id: str,
         body: dict,
         user_data: Optional[dict],
+        target_compressed_count: Optional[int],
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ):
         """
         Generates summary asynchronously (runs in background, does not block response).
@@ -714,17 +1001,18 @@ class Filter:
         3. Generate summary for the remaining middle messages.
         """
         try:
-            if self.valves.debug_mode:
-                print(f"\n[🤖 Async Summary Task] Starting...")
+            await self._log(
+                f"\n[🤖 Async Summary Task] Starting...", event_call=__event_call__
+            )
 
             # 1. Get target compression progress
-            # Prioritize getting from temp_state (calculated by inlet). If unavailable (e.g., after restart), assume current is full history.
-            target_compressed_count = self.temp_state.pop(chat_id, None)
+            # If target_compressed_count is not passed (should not happen with new logic), estimate it
             if target_compressed_count is None:
                 target_compressed_count = max(0, len(messages) - self.valves.keep_last)
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 Async Summary Task] ⚠️ Could not get inlet state, estimating progress using current message count: {target_compressed_count}"
+                await self._log(
+                    f"[🤖 Async Summary Task] ⚠️ target_compressed_count is None, estimating: {target_compressed_count}",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
             # 2. Determine the range of messages to compress (Middle)
@@ -735,25 +1023,33 @@ class Filter:
 
             # Ensure indices are valid
             if start_index >= end_index:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 Async Summary Task] Middle messages empty (Start: {start_index}, End: {end_index}), skipping"
+                await self._log(
+                    f"[🤖 Async Summary Task] Middle messages empty (Start: {start_index}, End: {end_index}), skipping",
+                    event_call=__event_call__,
                 )
                 return
 
             middle_messages = messages[start_index:end_index]
 
-            if self.valves.debug_mode:
-                print(
-                    f"[🤖 Async Summary Task] Middle messages to process: {len(middle_messages)}"
+            await self._log(
+                f"[🤖 Async Summary Task] Middle messages to process: {len(middle_messages)}",
+                event_call=__event_call__,
             )
 
             # 3. Check Token limit and truncate (Max Context Truncation)
             # [Optimization] Use the summary model's (if any) threshold to decide how many middle messages can be processed
             # This allows using a long-window model (like gemini-flash) to compress history exceeding the current model's window
-            summary_model_id = self.valves.summary_model or body.get(
-                "model", "gpt-3.5-turbo"
+            summary_model_id = self._clean_model_id(
+                self.valves.summary_model
+            ) or self._clean_model_id(body.get("model"))
+
+            if not summary_model_id:
+                await self._log(
+                    "[🤖 Async Summary Task] ⚠️ Summary model does not exist, skipping compression",
+                    type="warning",
+                    event_call=__event_call__,
                 )
+                return
 
             thresholds = self._get_model_thresholds(summary_model_id)
             # Note: Using the summary model's max context limit here
@@ -761,21 +1057,25 @@ class Filter:
                 "max_context_tokens", self.valves.max_context_tokens
             )
 
-            if self.valves.debug_mode:
-                print(
-                    f"[🤖 Async Summary Task] Using max limit for model {summary_model_id}: {max_context_tokens} Tokens"
+            await self._log(
+                f"[🤖 Async Summary Task] Using max limit for model {summary_model_id}: {max_context_tokens} Tokens",
+                event_call=__event_call__,
             )
 
-            # Calculate current total Tokens (using summary model for counting)
-            total_tokens = await asyncio.to_thread(
-                self._calculate_messages_tokens, messages
+            # Calculate tokens for middle messages only (plus buffer for prompt)
+            # We only send middle_messages to the summary model, so we shouldn't count the full history against its limit.
+            middle_tokens = await asyncio.to_thread(
+                self._calculate_messages_tokens, middle_messages
             )
+            # Add buffer for prompt and output (approx 2000 tokens)
+            estimated_input_tokens = middle_tokens + 2000
 
-            if total_tokens > max_context_tokens:
-                excess_tokens = total_tokens - max_context_tokens
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 Async Summary Task] ⚠️ Total Tokens ({total_tokens}) exceed summary model limit ({max_context_tokens}), need to remove approx {excess_tokens} Tokens"
+            if estimated_input_tokens > max_context_tokens:
+                excess_tokens = estimated_input_tokens - max_context_tokens
+                await self._log(
+                    f"[🤖 Async Summary Task] ⚠️ Middle messages ({middle_tokens} Tokens) + Buffer exceed summary model limit ({max_context_tokens}), need to remove approx {excess_tokens} Tokens",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
                 # Remove from the head of middle_messages
@@ -784,19 +1084,21 @@ class Filter:
 
                 while removed_tokens < excess_tokens and middle_messages:
                     msg_to_remove = middle_messages.pop(0)
-                    msg_tokens = self._count_tokens(str(msg_to_remove.get("content", "")))
+                    msg_tokens = self._count_tokens(
+                        str(msg_to_remove.get("content", ""))
+                    )
                     removed_tokens += msg_tokens
                     removed_count += 1
 
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 Async Summary Task] Removed {removed_count} messages, totaling {removed_tokens} Tokens"
+                await self._log(
+                    f"[🤖 Async Summary Task] Removed {removed_count} messages, totaling {removed_tokens} Tokens",
+                    event_call=__event_call__,
                 )
 
             if not middle_messages:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 Async Summary Task] Middle messages empty after truncation, skipping summary generation"
+                await self._log(
+                    f"[🤖 Async Summary Task] Middle messages empty after truncation, skipping summary generation",
+                    event_call=__event_call__,
                 )
                 return
 
@@ -819,13 +1121,25 @@ class Filter:
                 )
 
             new_summary = await self._call_summary_llm(
-                None, conversation_text, body, user_data
+                None,
+                conversation_text,
+                {**body, "model": summary_model_id},
+                user_data,
+                __event_call__,
             )
 
+            if not new_summary:
+                await self._log(
+                    "[🤖 Async Summary Task] ⚠️ Summary generation returned empty result, skipping save",
+                    type="warning",
+                    event_call=__event_call__,
+                )
+                return
+
             # 6. Save new summary
-            if self.valves.debug_mode:
-                print(
-                    "[Optimization] Saving summary in a background thread to avoid blocking the event loop."
+            await self._log(
+                "[Optimization] Saving summary in a background thread to avoid blocking the event loop.",
+                event_call=__event_call__,
             )
 
             await asyncio.to_thread(
@@ -844,16 +1158,34 @@ class Filter:
                     }
                 )
 
-            if self.valves.debug_mode:
-                print(
-                    f"[🤖 Async Summary Task] ✅ Complete! New summary length: {len(new_summary)} characters"
+            await self._log(
+                f"[🤖 Async Summary Task] ✅ Complete! New summary length: {len(new_summary)} characters",
+                type="success",
+                event_call=__event_call__,
             )
-                print(
-                    f"[🤖 Async Summary Task] Progress update: Compressed up to original message {target_compressed_count}"
+            await self._log(
+                f"[🤖 Async Summary Task] Progress update: Compressed up to original message {target_compressed_count}",
+                event_call=__event_call__,
             )
 
         except Exception as e:
-            print(f"[🤖 Async Summary Task] ❌ Error: {str(e)}")
+            await self._log(
+                f"[🤖 Async Summary Task] ❌ Error: {str(e)}",
+                type="error",
+                event_call=__event_call__,
+            )
+
+            if __event_emitter__:
+                await __event_emitter__(
+                    {
+                        "type": "status",
+                        "data": {
+                            "description": f"Summary Error: {str(e)[:100]}...",
+                            "done": True,
+                        },
+                    }
+                )
+
             import traceback
 
             traceback.print_exc()
@@ -890,12 +1222,15 @@ class Filter:
         new_conversation_text: str,
         body: dict,
         user_data: dict,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> str:
         """
         Calls the LLM to generate a summary using Open WebUI's built-in method.
         """
-        if self.valves.debug_mode:
-            print(f"[🤖 LLM Call] Using Open WebUI's built-in method")
+        await self._log(
+            f"[🤖 LLM Call] Using Open WebUI's built-in method",
+            event_call=__event_call__,
+        )
 
         # Build summary prompt (Optimized)
         summary_prompt = f"""
@@ -932,10 +1267,19 @@ This conversation may contain previous summaries (as system messages or text) an
 Based on the content above, generate the summary:
 """
         # Determine the model to use
-        model = self.valves.summary_model or body.get("model", "")
+        model = self._clean_model_id(self.valves.summary_model) or self._clean_model_id(
+            body.get("model")
+        )
 
-        if self.valves.debug_mode:
-            print(f"[🤖 LLM Call] Model: {model}")
+        if not model:
+            await self._log(
+                "[🤖 LLM Call] ⚠️ Summary model does not exist, skipping summary generation",
+                type="warning",
+                event_call=__event_call__,
+            )
+            return ""
+
+        await self._log(f"[🤖 LLM Call] Model: {model}", event_call=__event_call__)
 
         # Build payload
         payload = {
@@ -953,18 +1297,19 @@ Based on the content above, generate the summary:
                 raise ValueError("Could not get user ID")
 
             # [Optimization] Get user object in a background thread to avoid blocking the event loop.
-            if self.valves.debug_mode:
-                print(
-                    "[Optimization] Getting user object in a background thread to avoid blocking the event loop."
+            await self._log(
+                "[Optimization] Getting user object in a background thread to avoid blocking the event loop.",
+                event_call=__event_call__,
             )
             user = await asyncio.to_thread(Users.get_user_by_id, user_id)
 
             if not user:
                 raise ValueError(f"Could not find user: {user_id}")
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM Call] User: {user.email}")
-                print(f"[🤖 LLM Call] Sending request...")
+            await self._log(
+                f"[🤖 LLM Call] User: {user.email}\n[🤖 LLM Call] Sending request...",
+                event_call=__event_call__,
+            )
 
             # Create Request object
             request = Request(scope={"type": "http", "app": webui_app})
@@ -977,20 +1322,31 @@ Based on the content above, generate the summary:
 
             summary = response["choices"][0]["message"]["content"].strip()
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM Call] ✅ Successfully received summary")
+            await self._log(
+                f"[🤖 LLM Call] ✅ Successfully received summary",
+                type="success",
+                event_call=__event_call__,
+            )
 
             return summary
 
         except Exception as e:
-            error_message = f"Error occurred while calling LLM ({model}) to generate summary: {str(e)}"
+            error_msg = str(e)
+            # Handle specific error messages
+            if "Model not found" in error_msg:
+                error_message = f"Summary model '{model}' not found."
+            else:
+                error_message = f"Summary LLM Error ({model}): {error_msg}"
             if not self.valves.summary_model:
                 error_message += (
                     "\n[Hint] You did not specify a summary_model, so the filter attempted to use the current conversation's model. "
                     "If this is a pipeline (Pipe) model or an incompatible model, please specify a compatible summary model (e.g., 'gemini-2.5-flash') in the configuration."
                 )
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM Call] ❌ {error_message}")
+            await self._log(
+                f"[🤖 LLM Call] ❌ {error_message}",
+                type="error",
+                event_call=__event_call__,
+            )
 
             raise Exception(error_message)

plugins/filters/async-context-compression/async_context_compression_cn.py

@@ -5,7 +5,8 @@ author: Fu-Jie
 author_url: https://github.com/Fu-Jie
 funding_url: https://github.com/Fu-Jie/awesome-openwebui
 description: 通过智能摘要和消息压缩，降低长对话的 token 消耗，同时保持对话连贯性。
-version: 1.1.0
+version: 1.1.3
+openwebui_id: 5c0617cb-a9e4-4bd6-a440-d276534ebd18
 license: MIT
 
 ═══════════════════════════════════════════════════════════════════════════════
@@ -137,6 +138,10 @@ debug_mode (调试模式)
   默认: true
   说明: 在日志中打印详细的调试信息。生产环境建议设为 `false`。
 
+show_debug_log (前端调试日志)
+  默认: false
+  说明: 在浏览器控制台打印调试日志 (F12)。便于前端调试。
+
 🔧 部署配置
 ═══════════════════════════════════════════════════════
 
@@ -285,7 +290,8 @@ class Filter:
         self.valves = self.Valves()
         self._db_engine = owui_engine
         self._SessionLocal = owui_Session
-        self.temp_state = {}  # 用于在 inlet 和 outlet 之间传递临时数据
+        self._SessionLocal = owui_Session
+        self._init_database()
         self._init_database()
 
     def _init_database(self):
@@ -344,6 +350,9 @@ class Filter:
             default=0.1, ge=0.0, le=2.0, description="摘要生成的温度参数"
         )
         debug_mode: bool = Field(default=True, description="调试模式，打印详细日志")
+        show_debug_log: bool = Field(
+            default=False, description="在浏览器控制台打印调试日志 (F12)"
+        )
 
     def _save_summary(self, chat_id: str, summary: str, compressed_count: int):
         """保存摘要到数据库"""
@@ -425,9 +434,7 @@ class Filter:
         # 回退策略：粗略估算 (1 token ≈ 4 chars)
         return len(text) // 4
 
-    def _calculate_messages_tokens(
-        self, messages: List[Dict]
-    ) -> int:
+    def _calculate_messages_tokens(self, messages: List[Dict]) -> int:
         """计算消息列表的总 Token 数"""
         total_tokens = 0
         for msg in messages:
@@ -465,41 +472,101 @@ class Filter:
             "max_context_tokens": self.valves.max_context_tokens,
         }
 
-    def _inject_summary_to_first_message(self, message: dict, summary: str) -> dict:
-        """将摘要注入到第一条消息中（追加到内容前面）"""
-        content = message.get("content", "")
-        summary_block = f"【历史对话摘要】\n{summary}\n\n---\n以下是最近的对话：\n\n"
-
-        # 处理不同内容类型
-        if isinstance(content, list):  # 多模态内容
-            # 查找第一个文本部分并在其前面插入摘要
-            new_content = []
-            summary_inserted = False
-
-            for part in content:
-                if (
-                    isinstance(part, dict)
-                    and part.get("type") == "text"
-                    and not summary_inserted
+    async def _emit_debug_log(
+        self,
+        __event_call__,
+        chat_id: str,
+        original_count: int,
+        compressed_count: int,
+        summary_length: int,
+        kept_first: int,
+        kept_last: int,
     ):
-                    # 在第一个文本部分前插入摘要
-                    new_content.append(
-                        {"type": "text", "text": summary_block + part.get("text", "")}
+        """Emit debug log to browser console via JS execution"""
+        if not self.valves.show_debug_log or not __event_call__:
+            return
+
+        try:
+            # Prepare data for JS
+            log_data = {
+                "chatId": chat_id,
+                "originalCount": original_count,
+                "compressedCount": compressed_count,
+                "summaryLength": summary_length,
+                "keptFirst": kept_first,
+                "keptLast": kept_last,
+                "ratio": (
+                    f"{(1 - compressed_count/original_count)*100:.1f}%"
+                    if original_count > 0
+                    else "0%"
+                ),
+            }
+
+            # Construct JS code
+            js_code = f"""
+                (async function() {{
+                    console.group("🗜️ Async Context Compression Debug");
+                    console.log("Chat ID:", {json.dumps(chat_id)});
+                    console.log("Messages:", {original_count} + " -> " + {compressed_count});
+                    console.log("Compression Ratio:", {json.dumps(log_data['ratio'])});
+                    console.log("Summary Length:", {summary_length} + " chars");
+                    console.log("Configuration:", {{
+                        "Keep First": {kept_first},
+                        "Keep Last": {kept_last}
+                    }});
+                    console.groupEnd();
+                }})();
+            """
+
+            await __event_call__(
+                {
+                    "type": "execute",
+                    "data": {"code": js_code},
+                }
             )
-                    summary_inserted = True
-                else:
-                    new_content.append(part)
+        except Exception as e:
+            print(f"Error emitting debug log: {e}")
 
-            # 如果没有文本部分，在开头插入
-            if not summary_inserted:
-                new_content.insert(0, {"type": "text", "text": summary_block})
+    async def _log(self, message: str, type: str = "info", event_call=None):
+        """统一日志输出到后端 (print) 和前端 (console.log)"""
+        # 后端日志
+        if self.valves.debug_mode:
+            print(message)
 
-            message["content"] = new_content
+        # 前端日志
+        if self.valves.show_debug_log and event_call:
+            try:
+                css = "color: #3b82f6;"  # 默认蓝色
+                if type == "error":
+                    css = "color: #ef4444; font-weight: bold;"  # 红色
+                elif type == "warning":
+                    css = "color: #f59e0b;"  # 橙色
+                elif type == "success":
+                    css = "color: #10b981; font-weight: bold;"  # 绿色
 
-        elif isinstance(content, str):  # 纯文本
-            message["content"] = summary_block + content
+                # 清理前端消息：移除分隔符和多余换行
+                lines = message.split("\n")
+                # 保留不以大量等号或连字符开头的行
+                filtered_lines = [
+                    line
+                    for line in lines
+                    if not line.strip().startswith("====")
+                    and not line.strip().startswith("----")
+                ]
+                clean_message = "\n".join(filtered_lines).strip()
 
-        return message
+                if not clean_message:
+                    return
+
+                # 转义消息中的引号和换行符
+                safe_message = clean_message.replace('"', '\\"').replace("\n", "\\n")
+
+                js_code = f"""
+                    console.log("%c[压缩] {safe_message}", "{css}");
+                """
+                await event_call({"type": "execute", "data": {"code": js_code}})
+            except Exception as e:
+                print(f"发送前端日志失败: {e}")
 
     async def inlet(
         self,
@@ -507,6 +574,7 @@ class Filter:
         __user__: Optional[dict] = None,
         __metadata__: dict = None,
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> dict:
         """
         在发送到 LLM 之前执行
@@ -515,24 +583,24 @@ class Filter:
         messages = body.get("messages", [])
         chat_id = __metadata__["chat_id"]
 
-        if self.valves.debug_mode:
-            print(f"\n{'='*60}")
-            print(f"[Inlet] Chat ID: {chat_id}")
-            print(f"[Inlet] 收到 {len(messages)} 条消息")
+        if self.valves.debug_mode or self.valves.show_debug_log:
+            await self._log(
+                f"\n{'='*60}\n[Inlet] Chat ID: {chat_id}\n[Inlet] 收到 {len(messages)} 条消息",
+                event_call=__event_call__,
+            )
 
         # 记录原始消息的目标压缩进度，供 outlet 使用
         # 目标是压缩到倒数第 keep_last 条之前
         target_compressed_count = max(0, len(messages) - self.valves.keep_last)
 
-        # [优化] 简单的状态清理检查
-        if chat_id in self.temp_state:
-            if self.valves.debug_mode:
-                print(f"[Inlet] ⚠️ 覆盖未消费的旧状态 (Chat ID: {chat_id})")
+        # 记录原始消息的目标压缩进度，供 outlet 使用
+        # 目标是压缩到倒数第 keep_last 条之前
+        target_compressed_count = max(0, len(messages) - self.valves.keep_last)
 
-        self.temp_state[chat_id] = target_compressed_count
-
-        if self.valves.debug_mode:
-            print(f"[Inlet] 记录目标压缩进度: {target_compressed_count}")
+        await self._log(
+            f"[Inlet] 记录目标压缩进度: {target_compressed_count}",
+            event_call=__event_call__,
+        )
 
         # 加载摘要记录
         summary_record = await asyncio.to_thread(self._load_summary_record, chat_id)
@@ -560,7 +628,7 @@ class Filter:
                 f"---\n"
                 f"以下是最近的对话："
             )
-            summary_msg = {"role": "user", "content": summary_content}
+            summary_msg = {"role": "assistant", "content": summary_content}
 
             # 3. 尾部消息 (Tail) - 从上次压缩点开始的所有消息
             # 注意：这里必须确保不重复包含头部消息
@@ -581,9 +649,21 @@ class Filter:
                     }
                 )
 
-            if self.valves.debug_mode:
-                print(
-                    f"[Inlet] 应用摘要: Head({len(head_messages)}) + Summary + Tail({len(tail_messages)})"
+            await self._log(
+                f"[Inlet] 应用摘要: Head({len(head_messages)}) + Summary + Tail({len(tail_messages)})",
+                type="success",
+                event_call=__event_call__,
+            )
+
+            # Emit debug log to frontend (Keep the structured log as well)
+            await self._emit_debug_log(
+                __event_call__,
+                chat_id,
+                len(messages),
+                len(final_messages),
+                len(summary_record.summary),
+                self.valves.keep_first,
+                self.valves.keep_last,
             )
         else:
             # 没有摘要，使用原始消息
@@ -591,9 +671,10 @@ class Filter:
 
         body["messages"] = final_messages
 
-        if self.valves.debug_mode:
-            print(f"[Inlet] 最终发送: {len(body['messages'])} 条消息")
-            print(f"{'='*60}\n")
+        await self._log(
+            f"[Inlet] 最终发送: {len(body['messages'])} 条消息\n{'='*60}\n",
+            event_call=__event_call__,
+        )
 
         return body
 
@@ -603,29 +684,43 @@ class Filter:
         __user__: Optional[dict] = None,
         __metadata__: dict = None,
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> dict:
         """
         在 LLM 响应完成后执行
         在后台计算 Token 数并触发摘要生成（不阻塞当前响应，不影响内容输出）
         """
         chat_id = __metadata__["chat_id"]
-        model = body.get("model", "gpt-3.5-turbo")
+        model = body.get("model") or ""
 
-        if self.valves.debug_mode:
-            print(f"\n{'='*60}")
-            print(f"[Outlet] Chat ID: {chat_id}")
-            print(f"[Outlet] 响应完成")
+        # 直接计算目标压缩进度
+        # 假设 outlet 中的 body['messages'] 包含完整历史（包括新响应）
+        messages = body.get("messages", [])
+        target_compressed_count = max(0, len(messages) - self.valves.keep_last)
+
+        if self.valves.debug_mode or self.valves.show_debug_log:
+            await self._log(
+                f"\n{'='*60}\n[Outlet] Chat ID: {chat_id}\n[Outlet] 响应完成\n[Outlet] 计算目标压缩进度: {target_compressed_count} (消息数: {len(messages)})",
+                event_call=__event_call__,
+            )
 
         # 在后台异步处理 Token 计算和摘要生成（不等待完成，不影响输出）
         asyncio.create_task(
             self._check_and_generate_summary_async(
-                chat_id, model, body, __user__, __event_emitter__
+                chat_id,
+                model,
+                body,
+                __user__,
+                target_compressed_count,
+                __event_emitter__,
+                __event_call__,
             )
         )
 
-        if self.valves.debug_mode:
-            print(f"[Outlet] 后台处理已启动")
-            print(f"{'='*60}\n")
+        await self._log(
+            f"[Outlet] 后台处理已启动\n{'='*60}\n",
+            event_call=__event_call__,
+        )
 
         return body
 
@@ -635,7 +730,9 @@ class Filter:
         model: str,
         body: dict,
         user_data: Optional[dict],
+        target_compressed_count: Optional[int],
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ):
         """
         后台处理：计算 Token 数并生成摘要（不阻塞响应）
@@ -649,36 +746,58 @@ class Filter:
                 "compression_threshold_tokens", self.valves.compression_threshold_tokens
             )
 
-            if self.valves.debug_mode:
-                print(f"\n[🔍 后台计算] 开始 Token 计数...")
+            await self._log(
+                f"\n[🔍 后台计算] 开始 Token 计数...",
+                event_call=__event_call__,
+            )
 
             # 在后台线程中计算 Token 数
             current_tokens = await asyncio.to_thread(
                 self._calculate_messages_tokens, messages
             )
 
-            if self.valves.debug_mode:
-                print(f"[🔍 后台计算] Token 数: {current_tokens}")
+            await self._log(
+                f"[🔍 后台计算] Token 数: {current_tokens}",
+                event_call=__event_call__,
+            )
 
             # 检查是否需要压缩
             if current_tokens >= compression_threshold_tokens:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🔍 后台计算] ⚡ 触发压缩阈值 (Token: {current_tokens} >= {compression_threshold_tokens})"
+                await self._log(
+                    f"[🔍 后台计算] ⚡ 触发压缩阈值 (Token: {current_tokens} >= {compression_threshold_tokens})",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
                 # 继续生成摘要
                 await self._generate_summary_async(
-                    messages, chat_id, body, user_data, __event_emitter__
+                    messages,
+                    chat_id,
+                    body,
+                    user_data,
+                    target_compressed_count,
+                    __event_emitter__,
+                    __event_call__,
                 )
             else:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🔍 后台计算] 未触发压缩阈值 (Token: {current_tokens} < {compression_threshold_tokens})"
+                await self._log(
+                    f"[🔍 后台计算] 未触发压缩阈值 (Token: {current_tokens} < {compression_threshold_tokens})",
+                    event_call=__event_call__,
                 )
 
         except Exception as e:
-            print(f"[🔍 后台计算] ❌ 错误: {str(e)}")
+            await self._log(
+                f"[🔍 后台计算] ❌ 错误: {str(e)}",
+                type="error",
+                event_call=__event_call__,
+            )
+
+    def _clean_model_id(self, model_id: Optional[str]) -> Optional[str]:
+        """Cleans the model ID by removing whitespace and quotes."""
+        if not model_id:
+            return None
+        cleaned = model_id.strip().strip('"').strip("'")
+        return cleaned if cleaned else None
 
     async def _generate_summary_async(
         self,
@@ -686,7 +805,9 @@ class Filter:
         chat_id: str,
         body: dict,
         user_data: Optional[dict],
+        target_compressed_count: Optional[int],
         __event_emitter__: Callable[[Any], Awaitable[None]] = None,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ):
         """
         异步生成摘要（后台执行，不阻塞响应）
@@ -696,17 +817,16 @@ class Filter:
         3. 对剩余的中间消息生成摘要。
         """
         try:
-            if self.valves.debug_mode:
-                print(f"\n[🤖 异步摘要任务] 开始...")
+            await self._log(f"\n[🤖 异步摘要任务] 开始...", event_call=__event_call__)
 
             # 1. 获取目标压缩进度
-            # 优先从 temp_state 获取（由 inlet 计算），如果获取不到（例如重启后），则假设当前是完整历史
-            target_compressed_count = self.temp_state.pop(chat_id, None)
+            # 如果未传递 target_compressed_count（新逻辑下不应发生），则进行估算
             if target_compressed_count is None:
                 target_compressed_count = max(0, len(messages) - self.valves.keep_last)
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 异步摘要任务] ⚠️ 无法获取 inlet 状态，使用当前消息数估算进度: {target_compressed_count}"
+                await self._log(
+                    f"[🤖 异步摘要任务] ⚠️ target_compressed_count 为 None，进行估算: {target_compressed_count}",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
             # 2. 确定待压缩的消息范围 (Middle)
@@ -717,21 +837,33 @@ class Filter:
 
             # 确保索引有效
             if start_index >= end_index:
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 异步摘要任务] 中间消息为空 (Start: {start_index}, End: {end_index})，跳过"
+                await self._log(
+                    f"[🤖 异步摘要任务] 中间消息为空 (Start: {start_index}, End: {end_index})，跳过",
+                    event_call=__event_call__,
                 )
                 return
 
             middle_messages = messages[start_index:end_index]
 
-            if self.valves.debug_mode:
-                print(f"[🤖 异步摘要任务] 待处理中间消息: {len(middle_messages)} 条")
+            await self._log(
+                f"[🤖 异步摘要任务] 待处理中间消息: {len(middle_messages)} 条",
+                event_call=__event_call__,
+            )
 
             # 3. 检查 Token 上限并截断 (Max Context Truncation)
             # [优化] 使用摘要模型(如果有)的阈值来决定能处理多少中间消息
             # 这样可以用长窗口模型(如 gemini-flash)来压缩超过当前模型窗口的历史记录
-            summary_model_id = self.valves.summary_model or body.get("model")
+            summary_model_id = self._clean_model_id(
+                self.valves.summary_model
+            ) or self._clean_model_id(body.get("model"))
+
+            if not summary_model_id:
+                await self._log(
+                    "[🤖 异步摘要任务] ⚠️ 摘要模型不存在，跳过压缩",
+                    type="warning",
+                    event_call=__event_call__,
+                )
+                return
 
             thresholds = self._get_model_thresholds(summary_model_id)
             # 注意：这里使用的是摘要模型的最大上下文限制
@@ -739,21 +871,25 @@ class Filter:
                 "max_context_tokens", self.valves.max_context_tokens
             )
 
-            if self.valves.debug_mode:
-                print(
-                    f"[🤖 异步摘要任务] 使用模型 {summary_model_id} 的上限: {max_context_tokens} Tokens"
+            await self._log(
+                f"[🤖 异步摘要任务] 使用模型 {summary_model_id} 的上限: {max_context_tokens} Tokens",
+                event_call=__event_call__,
             )
 
-            # 计算当前总 Token (使用摘要模型进行计数)
-            total_tokens = await asyncio.to_thread(
-                self._calculate_messages_tokens, messages
+            # 计算中间消息的 Token (加上提示词的缓冲)
+            # 我们只把 middle_messages 发送给摘要模型，所以不应该把完整历史计入限制
+            middle_tokens = await asyncio.to_thread(
+                self._calculate_messages_tokens, middle_messages
             )
+            # 增加提示词和输出的缓冲 (约 2000 Tokens)
+            estimated_input_tokens = middle_tokens + 2000
 
-            if total_tokens > max_context_tokens:
-                excess_tokens = total_tokens - max_context_tokens
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 异步摘要任务] ⚠️ 总 Token ({total_tokens}) 超过摘要模型上限 ({max_context_tokens})，需要移除约 {excess_tokens} Token"
+            if estimated_input_tokens > max_context_tokens:
+                excess_tokens = estimated_input_tokens - max_context_tokens
+                await self._log(
+                    f"[🤖 异步摘要任务] ⚠️ 中间消息 ({middle_tokens} Tokens) + 缓冲超过摘要模型上限 ({max_context_tokens})，需要移除约 {excess_tokens} Token",
+                    type="warning",
+                    event_call=__event_call__,
                 )
 
                 # 从 middle_messages 头部开始移除
@@ -768,14 +904,16 @@ class Filter:
                     removed_tokens += msg_tokens
                     removed_count += 1
 
-                if self.valves.debug_mode:
-                    print(
-                        f"[🤖 异步摘要任务] 已移除 {removed_count} 条消息，共 {removed_tokens} Token"
+                await self._log(
+                    f"[🤖 异步摘要任务] 已移除 {removed_count} 条消息，共 {removed_tokens} Token",
+                    event_call=__event_call__,
                 )
 
             if not middle_messages:
-                if self.valves.debug_mode:
-                    print(f"[🤖 异步摘要任务] 截断后中间消息为空，跳过摘要生成")
+                await self._log(
+                    f"[🤖 异步摘要任务] 截断后中间消息为空，跳过摘要生成",
+                    event_call=__event_call__,
+                )
                 return
 
             # 4. 构建对话文本
@@ -797,12 +935,26 @@ class Filter:
                 )
 
             new_summary = await self._call_summary_llm(
-                None, conversation_text, body, user_data
+                None,
+                conversation_text,
+                {**body, "model": summary_model_id},
+                user_data,
+                __event_call__,
             )
 
+            if not new_summary:
+                await self._log(
+                    "[🤖 异步摘要任务] ⚠️ 摘要生成返回空结果，跳过保存",
+                    type="warning",
+                    event_call=__event_call__,
+                )
+                return
+
             # 6. 保存新摘要
-            if self.valves.debug_mode:
-                print("[优化] 正在后台线程中保存摘要，以避免阻塞事件循环。")
+            await self._log(
+                "[优化] 在后台线程中保存摘要以避免阻塞事件循环。",
+                event_call=__event_call__,
+            )
 
             await asyncio.to_thread(
                 self._save_summary, chat_id, new_summary, target_compressed_count
@@ -814,32 +966,52 @@ class Filter:
                     {
                         "type": "status",
                         "data": {
-                            "description": f"上下文摘要已更新 (已压缩 {len(middle_messages)} 条消息)",
+                            "description": f"上下文摘要已更新 (压缩了 {len(middle_messages)} 条消息)",
                             "done": True,
                         },
                     }
                 )
 
-            if self.valves.debug_mode:
-                print(f"[🤖 异步摘要任务] ✅ 完成！新摘要长度: {len(new_summary)} 字符")
-                print(
-                    f"[🤖 异步摘要任务] 进度更新: 已压缩至原始第 {target_compressed_count} 条消息"
+            await self._log(
+                f"[🤖 异步摘要任务] ✅ 完成！新摘要长度: {len(new_summary)} 字符",
+                type="success",
+                event_call=__event_call__,
+            )
+            await self._log(
+                f"[🤖 异步摘要任务] 进度更新: 已压缩至原始消息 {target_compressed_count}",
+                event_call=__event_call__,
             )
 
         except Exception as e:
-            print(f"[🤖 异步摘要任务] ❌ 错误: {str(e)}")
+            await self._log(
+                f"[🤖 异步摘要任务] ❌ 错误: {str(e)}",
+                type="error",
+                event_call=__event_call__,
+            )
+
+            if __event_emitter__:
+                await __event_emitter__(
+                    {
+                        "type": "status",
+                        "data": {
+                            "description": f"摘要生成错误: {str(e)[:100]}...",
+                            "done": True,
+                        },
+                    }
+                )
+
             import traceback
 
             traceback.print_exc()
 
     def _format_messages_for_summary(self, messages: list) -> str:
-        """格式化消息用于摘要"""
+        """Formats messages for summarization."""
         formatted = []
         for i, msg in enumerate(messages, 1):
             role = msg.get("role", "unknown")
             content = msg.get("content", "")
 
-            # 处理多模态内容
+            # Handle multimodal content
             if isinstance(content, list):
                 text_parts = []
                 for part in content:
@@ -847,10 +1019,10 @@ class Filter:
                         text_parts.append(part.get("text", ""))
                 content = " ".join(text_parts)
 
-            # 处理角色名称
-            role_name = {"user": "用户", "assistant": "助手"}.get(role, role)
+            # Handle role name
+            role_name = {"user": "User", "assistant": "Assistant"}.get(role, role)
 
-            # 限制每条消息的长度，避免过长
+            # Limit length of each message to avoid excessive length
             if len(content) > 500:
                 content = content[:500] + "..."
 
@@ -864,12 +1036,15 @@ class Filter:
         new_conversation_text: str,
         body: dict,
         user_data: dict,
+        __event_call__: Callable[[Any], Awaitable[None]] = None,
     ) -> str:
         """
-        使用 Open WebUI 内置方法调用 LLM 生成摘要
+        调用 LLM 生成摘要，使用 Open Web UI 的内置方法。
         """
-        if self.valves.debug_mode:
-            print(f"[🤖 LLM 调用] 使用 Open WebUI 内置方法")
+        await self._log(
+            f"[🤖 LLM 调用] 使用 Open Web UI 内置方法",
+            event_call=__event_call__,
+        )
 
         # 构建摘要提示词 (优化版)
         summary_prompt = f"""
@@ -906,10 +1081,19 @@ class Filter:
 请根据上述内容，生成摘要：
 """
         # 确定使用的模型
-        model = self.valves.summary_model or body.get("model", "")
+        model = self._clean_model_id(self.valves.summary_model) or self._clean_model_id(
+            body.get("model")
+        )
 
-        if self.valves.debug_mode:
-            print(f"[🤖 LLM 调用] 模型: {model}")
+        if not model:
+            await self._log(
+                "[🤖 LLM 调用] ⚠️ 摘要模型不存在，跳过摘要生成",
+                type="warning",
+                event_call=__event_call__,
+            )
+            return ""
+
+        await self._log(f"[🤖 LLM 调用] 模型: {model}", event_call=__event_call__)
 
         # 构建 payload
         payload = {
@@ -926,17 +1110,20 @@ class Filter:
             if not user_id:
                 raise ValueError("无法获取用户 ID")
 
-            # [优化] 在后台线程中获取用户对象，以避免阻塞事件循环
-            if self.valves.debug_mode:
-                print("[优化] 正在后台线程中获取用户对象，以避免阻塞事件循环。")
+            # [优化] 在后台线程中获取用户对象以避免阻塞事件循环
+            await self._log(
+                "[优化] 在后台线程中获取用户对象以避免阻塞事件循环。",
+                event_call=__event_call__,
+            )
             user = await asyncio.to_thread(Users.get_user_by_id, user_id)
 
             if not user:
                 raise ValueError(f"无法找到用户: {user_id}")
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM 调用] 用户: {user.email}")
-                print(f"[🤖 LLM 调用] 发送请求...")
+            await self._log(
+                f"[🤖 LLM 调用] 用户: {user.email}\n[🤖 LLM 调用] 发送请求...",
+                event_call=__event_call__,
+            )
 
             # 创建 Request 对象
             request = Request(scope={"type": "http", "app": webui_app})
@@ -949,20 +1136,31 @@ class Filter:
 
             summary = response["choices"][0]["message"]["content"].strip()
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM 调用] ✅ 成功获取摘要")
+            await self._log(
+                f"[🤖 LLM 调用] ✅ 成功接收摘要",
+                type="success",
+                event_call=__event_call__,
+            )
 
             return summary
 
         except Exception as e:
-            error_message = f"调用 LLM ({model}) 生成摘要时发生错误: {str(e)}"
+            error_msg = str(e)
+            # Handle specific error messages
+            if "Model not found" in error_msg:
+                error_message = f"摘要模型 '{model}' 不存在。"
+            else:
+                error_message = f"摘要 LLM 错误 ({model}): {error_msg}"
             if not self.valves.summary_model:
                 error_message += (
-                    "\n[提示] 您没有指定摘要模型 (summary_model)，因此尝试使用当前对话的模型。"
-                    "如果这是一个流水线（Pipe）模型或不兼容的模型，请在配置中指定一个兼容的摘要模型（如 'gemini-2.5-flash'）。"
+                    "\n[提示] 您未指定 summary_model，因此过滤器尝试使用当前对话的模型。"
+                    "如果这是流水线 (Pipe) 模型或不兼容的模型，请在配置中指定兼容的摘要模型 (例如 'gemini-2.5-flash')。"
                 )
 
-            if self.valves.debug_mode:
-                print(f"[🤖 LLM 调用] ❌ {error_message}")
+            await self._log(
+                f"[🤖 LLM 调用] ❌ {error_message}",
+                type="error",
+                event_call=__event_call__,
+            )
 
             raise Exception(error_message)

plugins/filters/context_enhancement_filter/context_enhancement_filter.py

@@ -1,12 +1,9 @@
 """
 title: Context & Model Enhancement Filter
-author: Fu-Jie
-author_url: https://github.com/Fu-Jie
-funding_url: https://github.com/Fu-Jie/awesome-openwebui
-version: 0.2
+version: 0.3
 
 description:
-    一个功能全面的 Filter 插件，用于增强请求上下文和优化模型功能。提供四大核心功能：
+    一个专注于增强请求上下文和优化模型功能的 Filter 插件。提供三大核心功能：
 
     1. 环境变量注入：在每条用户消息前自动注入用户环境变量（用户名、时间、时区、语言等）
        - 支持纯文本、图片、多模态消息
@@ -24,222 +21,24 @@ description:
        - 动态模型重定向
        - 智能化的模型识别和适配
 
-    4. 智能内容规范化：生产级的内容清洗与修复系统
-       - 智能修复损坏的代码块（前缀、后缀、缩进）
-       - 规范化 LaTeX 公式格式（行内/块级）
-       - 优化思维链标签（</thought>）格式
-       - 自动闭合未结束的代码块
-       - 智能列表格式修复
-       - 清理冗余的 XML 标签
-       - 可配置的规则系统
-
 features:
     - 自动化环境变量管理
     - 智能模型功能适配
     - 异步状态反馈
     - 幂等性保证
     - 多模型支持
-    - 智能内容清洗与规范化
 """
 
 from pydantic import BaseModel, Field
-from typing import Optional, List, Callable
+from typing import Optional
 import re
 import logging
-from dataclasses import dataclass, field
+import asyncio
 
 
 # 配置日志
 logger = logging.getLogger(__name__)
 
-@dataclass
-class NormalizerConfig:
-    """规范化配置类,用于动态启用/禁用特定规则"""
-    enable_escape_fix: bool = True          # 修复转义字符
-    enable_thought_tag_fix: bool = True     # 修复思考链标签
-    enable_code_block_fix: bool = True      # 修复代码块格式
-    enable_latex_fix: bool = True           # 修复 LaTeX 公式格式
-    enable_list_fix: bool = False            # 修复列表换行
-    enable_unclosed_block_fix: bool = True  # 修复未闭合代码块
-    enable_fullwidth_symbol_fix: bool = False # 修复代码内的全角符号
-    enable_xml_tag_cleanup: bool = True     # 清理 XML 残留标签
-    
-    # 自定义清理函数列表（高级扩展用）
-    custom_cleaners: List[Callable[[str], str]] = field(default_factory=list)
-
-class ContentNormalizer:
-    """LLM 输出内容规范化器 - 生产级实现"""
-    
-    # --- 1. 预编译正则表达式（性能优化） ---
-    _PATTERNS = {
-        # 代码块前缀：如果 ``` 前面不是行首也不是换行符
-        'code_block_prefix': re.compile(r'(?<!^)(?<!\n)(```)', re.MULTILINE),
-        
-        # 代码块后缀：匹配 ```语言名 后面紧跟非空白字符(没有换行)
-        # 匹配 ```python code 这种情况，但不匹配 ```python 或 ```python\n
-        'code_block_suffix': re.compile(r'(```[\w\+\-\.]*)[ \t]+([^\n\r])'),
-        
-        # 代码块缩进：行首的空白字符 + ```
-        'code_block_indent': re.compile(r'^[ \t]+(```)', re.MULTILINE),
-        
-        # 思考链标签：</thought> 后可能跟空格或换行
-        'thought_tag': re.compile(r'</thought>[ \t]*\n*'),
-        
-        # LaTeX 块级公式：\[ ... \]
-        'latex_bracket_block': re.compile(r'\\\[(.+?)\\\]', re.DOTALL),
-        # LaTeX 行内公式：\( ... \)
-        'latex_paren_inline': re.compile(r'\\\((.+?)\\\)'),
-        
-        # 列表项：非换行符 + 数字 + 点 + 空格 (e.g. "Text1. Item")
-        'list_item': re.compile(r'([^\n])(\d+\. )'),
-        
-        # XML 残留标签 (如 Claude 的 artifacts)
-        'xml_artifacts': re.compile(r'</?(?:antArtifact|antThinking|artifact)[^>]*>', re.IGNORECASE),
-    }
-    
-    def __init__(self, config: Optional[NormalizerConfig] = None):
-        self.config = config or NormalizerConfig()
-        self.applied_fixes = []
-    
-    def normalize(self, content: str) -> str:
-        """主入口：按顺序应用所有规范化规则"""
-        self.applied_fixes = []
-        if not content:
-            return content
-        
-        try:
-            # 1. 转义字符修复（必须最先执行，否则影响后续正则）
-            if self.config.enable_escape_fix:
-                original = content
-                content = self._fix_escape_characters(content)
-                if content != original:
-                    self.applied_fixes.append("修复转义字符")
-            
-            # 2. 思考链标签规范化
-            if self.config.enable_thought_tag_fix:
-                original = content
-                content = self._fix_thought_tags(content)
-                if content != original:
-                    self.applied_fixes.append("规范化思考链")
-            
-            # 3. 代码块格式修复
-            if self.config.enable_code_block_fix:
-                original = content
-                content = self._fix_code_blocks(content)
-                if content != original:
-                    self.applied_fixes.append("修复代码块格式")
-            
-            # 4. LaTeX 公式规范化
-            if self.config.enable_latex_fix:
-                original = content
-                content = self._fix_latex_formulas(content)
-                if content != original:
-                    self.applied_fixes.append("规范化 LaTeX 公式")
-            
-            # 5. 列表格式修复
-            if self.config.enable_list_fix:
-                original = content
-                content = self._fix_list_formatting(content)
-                if content != original:
-                    self.applied_fixes.append("修复列表格式")
-            
-            # 6. 未闭合代码块检测与修复
-            if self.config.enable_unclosed_block_fix:
-                original = content
-                content = self._fix_unclosed_code_blocks(content)
-                if content != original:
-                    self.applied_fixes.append("闭合未结束代码块")
-            
-            # 7. 全角符号转半角（仅代码块内）
-            if self.config.enable_fullwidth_symbol_fix:
-                original = content
-                content = self._fix_fullwidth_symbols_in_code(content)
-                if content != original:
-                    self.applied_fixes.append("全角符号转半角")
-            
-            # 8. XML 标签残留清理
-            if self.config.enable_xml_tag_cleanup:
-                original = content
-                content = self._cleanup_xml_tags(content)
-                if content != original:
-                    self.applied_fixes.append("清理 XML 标签")
-            
-            # 9. 执行自定义清理函数
-            for cleaner in self.config.custom_cleaners:
-                original = content
-                content = cleaner(content)
-                if content != original:
-                    self.applied_fixes.append("执行自定义清理")
-            
-            return content
-            
-        except Exception as e:
-            # 生产环境保底机制：如果清洗过程报错，返回原始内容，避免阻断服务
-            logger.error(f"内容规范化失败: {e}", exc_info=True)
-            return content
-    
-    def _fix_escape_characters(self, content: str) -> str:
-        """修复过度转义的字符"""
-        # 注意：先处理具体的转义序列，再处理通用的双反斜杠
-        content = content.replace("\\r\\n", "\n")
-        content = content.replace("\\n", "\n")
-        content = content.replace("\\t", "\t")
-        # 修复过度转义的反斜杠 (例如路径 C:\\Users)
-        content = content.replace("\\\\", "\\")
-        return content
-    
-    def _fix_thought_tags(self, content: str) -> str:
-        """规范化 </thought> 标签，统一为空两行"""
-        return self._PATTERNS['thought_tag'].sub("</thought>\n\n", content)
-    
-    def _fix_code_blocks(self, content: str) -> str:
-        """修复代码块格式（独占行、换行、去缩进）"""
-        # C: 移除代码块前的缩进（必须先执行，否则影响下面的判断）
-        content = self._PATTERNS['code_block_indent'].sub(r"\1", content)
-        # A: 确保 ``` 前有换行
-        content = self._PATTERNS['code_block_prefix'].sub(r"\n\1", content)
-        # B: 确保 ```语言标识 后有换行
-        content = self._PATTERNS['code_block_suffix'].sub(r"\1\n\2", content)
-        return content
-    
-    def _fix_latex_formulas(self, content: str) -> str:
-        """规范化 LaTeX 公式：\[ -> $$ (块级), \( -> $ (行内)"""
-        content = self._PATTERNS['latex_bracket_block'].sub(r"$$\1$$", content)
-        content = self._PATTERNS['latex_paren_inline'].sub(r"$\1$", content)
-        return content
-    
-    def _fix_list_formatting(self, content: str) -> str:
-        """修复列表项缺少换行的问题 (如 'text1. item' -> 'text\\n1. item')"""
-        return self._PATTERNS['list_item'].sub(r"\1\n\2", content)
-    
-    def _fix_unclosed_code_blocks(self, content: str) -> str:
-        """检测并修复未闭合的代码块"""
-        if content.count("```") % 2 != 0:
-            logger.warning("检测到未闭合的代码块，自动补全")
-            content += "\n```"
-        return content
-    
-    def _fix_fullwidth_symbols_in_code(self, content: str) -> str:
-        """在代码块内将全角符号转为半角（精细化操作）"""
-        # 常见误用的全角符号映射
-        FULLWIDTH_MAP = {
-            '，': ',', '。': '.', '（': '(', '）': ')',
-            '【': '[', '】': ']', '；': ';', '：': ':',
-            '？': '?', '！': '!', '"': '"', '"': '"',
-            ''': "'", ''': "'",
-        }
-        
-        parts = content.split("```")
-        # 代码块内容位于索引 1, 3, 5... (奇数位)
-        for i in range(1, len(parts), 2):
-            for full, half in FULLWIDTH_MAP.items():
-                parts[i] = parts[i].replace(full, half)
-        
-        return "```".join(parts)
-    
-    def _cleanup_xml_tags(self, content: str) -> str:
-        """移除无关的 XML 标签"""
-        return self._PATTERNS['xml_artifacts'].sub("", content)
 
 class Filter:
     class Valves(BaseModel):
@@ -349,13 +148,9 @@ class Filter:
                 body["model"] = body["model"] + "-search"
                 features["web_search"] = False
                 search_enabled_for_model = True
-            if user_email == "yi204o@qq.com":
-                features["web_search"] = False
 
         # 如果启用了模型本身的搜索能力，发送状态提示
         if search_enabled_for_model and __event_emitter__:
-            import asyncio
-
             try:
                 asyncio.create_task(
                     self._emit_search_status(__event_emitter__, model_name)
@@ -464,8 +259,6 @@ class Filter:
 
             # 环境变量注入成功后，发送状态提示给用户
             if env_injected and __event_emitter__:
-                import asyncio
-
                 try:
                     # 如果在异步环境中，使用 await
                     asyncio.create_task(self._emit_env_status(__event_emitter__))
@@ -506,67 +299,3 @@ class Filter:
             )
         except Exception as e:
             print(f"发送搜索状态提示时出错: {e}")
-
-    async def _emit_normalization_status(self, __event_emitter__, applied_fixes: List[str] = None):
-        """
-        发送内容规范化完成的状态提示
-        """
-        description = "✓ 内容已自动规范化"
-        if applied_fixes:
-            description += f"：{', '.join(applied_fixes)}"
-
-        try:
-            await __event_emitter__(
-                {
-                    "type": "status",
-                    "data": {
-                        "description": description,
-                        "done": True,
-                    },
-                }
-            )
-        except Exception as e:
-            print(f"发送规范化状态提示时出错: {e}")
-
-    def _contains_html(self, content: str) -> bool:
-        """
-        检测内容是否包含 HTML 标签
-        """
-        # 匹配常见的 HTML 标签
-        pattern = r"<\s*/?\s*(?:html|head|body|div|span|p|br|hr|ul|ol|li|table|thead|tbody|tfoot|tr|td|th|img|a|b|i|strong|em|code|pre|blockquote|h[1-6]|script|style|form|input|button|label|select|option|iframe|link|meta|title)\b"
-        return bool(re.search(pattern, content, re.IGNORECASE))
-
-    def outlet(self, body: dict, __user__: Optional[dict] = None, __event_emitter__=None) -> dict:
-        """
-        处理传出响应体，通过修改最后一条助手消息的内容。
-        使用 ContentNormalizer 进行全面的内容规范化。
-        """
-        if "messages" in body and body["messages"]:
-            last = body["messages"][-1]
-            content = last.get("content", "") or ""
-            
-            if last.get("role") == "assistant" and isinstance(content, str):
-                # 如果包含 HTML，跳过规范化，为了防止错误格式化
-                if self._contains_html(content):
-                    return body
-
-                # 初始化规范化器
-                normalizer = ContentNormalizer()
-                
-                # 执行规范化
-                new_content = normalizer.normalize(content)
-                
-                # 更新内容
-                if new_content != content:
-                    last["content"] = new_content
-                    # 如果内容发生了改变，发送状态提示
-                    if __event_emitter__:
-                        import asyncio
-                        try:
-                            # 传入 applied_fixes
-                            asyncio.create_task(self._emit_normalization_status(__event_emitter__, normalizer.applied_fixes))
-                        except RuntimeError:
-                            # 假如不在循环中，则忽略
-                            pass
-        
-        return body

plugins/filters/markdown_normalizer/FEATURES_CN.md

@@ -0,0 +1,162 @@
+# Markdown Normalizer 功能详解
+
+本插件旨在修复 LLM 输出中常见的 Markdown 格式问题，确保在 Open WebUI 中完美渲染。以下是支持的修复功能列表及示例。
+
+## 1. 代码块修复 (Code Block Fixes)
+
+### 1.1 去除代码块缩进
+LLM 有时会在代码块前添加空格缩进，导致渲染失效。本插件会自动移除这些缩进。
+
+**Before:**
+   ```python
+   print("hello")
+   ```
+
+**After:**
+```python
+print("hello")
+```
+
+### 1.2 补全代码块前后换行
+代码块标记 ` ``` ` 必须独占一行。如果 LLM 将其与文本混在一行，插件会自动修复。
+
+**Before:**
+Here is code:```python
+print("hello")```
+
+**After:**
+Here is code:
+```python
+print("hello")
+```
+
+### 1.3 修复语言标识符后的换行
+有时 LLM 会忘记在语言标识符（如 `python`）后换行。
+
+**Before:**
+```python print("hello")
+```
+
+**After:**
+```python
+print("hello")
+```
+
+### 1.4 自动闭合代码块
+如果输出被截断或 LLM 忘记闭合代码块，插件会自动添加结尾的 ` ``` `。
+
+**Before:**
+```python
+print("unfinished code...")
+
+**After:**
+```python
+print("unfinished code...")
+```
+
+## 2. LaTeX 公式规范化 (LaTeX Normalization)
+
+Open WebUI 使用 MathJax/KaTeX 渲染公式，通常需要 `$$` 或 `$` 包裹。本插件会将常见的 LaTeX 括号语法转换为标准格式。
+
+**Before:**
+块级公式：\[ E = mc^2 \]
+行内公式：\( a^2 + b^2 = c^2 \)
+
+**After:**
+块级公式：$$ E = mc^2 $$
+行内公式：$ a^2 + b^2 = c^2 $
+
+## 3. 转义字符清理 (Escape Character Fix)
+
+修复过度转义的字符，这常见于某些 API 返回的原始字符串中。
+
+**Before:**
+Line 1\\nLine 2\\tTabbed
+
+**After:**
+Line 1
+Line 2	Tabbed
+
+## 4. 思维链标签规范化 (Thought Tag Fix)
+**功能**: 
+1.  确保 `</thought>` 标签后有足够的空行，防止思维链内容与正文粘连。
+2.  **标准化标签**: 将 `<think>` (DeepSeek 等模型常用) 或 `<thinking>` 统一转换为 Open WebUI 标准的 `<thought>` 标签，以便正确触发 UI 的折叠功能。
+
+**默认**: 开启 (`enable_thought_tag_fix = True`)
+
+**示例**:
+*   **Before**: `<think>Thinking...</think>Response starts here.`
+*   **After**: 
+    ```xml
+    <thought>Thinking...</thought>
+
+    Response starts here.
+    ```
+
+## 5. 列表格式修复 (List Formatting Fix)
+
+*默认关闭，需在设置中开启*
+
+修复列表项缺少换行的问题。
+
+**Before:**
+Header1. Item 1
+
+**After:**
+Header
+1. Item 1
+
+## 6. 全角符号转半角 (Full-width Symbol Fix)
+
+*默认关闭，需在设置中开启*
+
+仅在**代码块内部**将全角符号转换为半角符号，防止代码因符号问题无法运行。
+
+**Before:**
+```python
+if x == 1：
+    print（"hello"）
+```
+
+**After:**
+```python
+if x == 1:
+    print("hello")
+```
+
+## 7. Mermaid 语法修复 (Mermaid Syntax Fix)
+**功能**: 修复 Mermaid 图表中常见的语法错误，特别是未加引号的标签包含特殊字符的情况。
+**默认**: 开启 (`enable_mermaid_fix = True`)
+**示例**:
+*   **Before**:
+    ```mermaid
+    graph TD
+    A[Label with (parens)] --> B(Label with [brackets])
+    ```
+*   **After**:
+    ```mermaid
+    graph TD
+    A["Label with (parens)"] --> B("Label with [brackets]")
+    ```
+
+## 8. XML 标签清理 (XML Cleanup)
+
+移除 LLM 输出中残留的无用 XML 标签（如 Claude 的 artifact 标签）。
+
+**Before:**
+Here is the result <antArtifact>hidden metadata</antArtifact>.
+
+**After:**
+## 9. 标题格式修复 (Heading Format Fix)
+**功能**: 修复标题标记 `#` 后缺少空格的问题。
+**默认**: 开启 (`enable_heading_fix = True`)
+**示例**:
+*   **Before**: `#Heading 1`
+*   **After**: `# Heading 1`
+
+## 10. 表格格式修复 (Table Format Fix)
+**功能**: 修复表格行末尾缺少管道符 `|` 的问题。
+**默认**: 开启 (`enable_table_fix = True`)
+**示例**:
+*   **Before**: `| Col 1 | Col 2`
+*   **After**: `| Col 1 | Col 2 |`

plugins/filters/markdown_normalizer/README.md

@@ -0,0 +1,46 @@
+# Markdown Normalizer Filter
+
+A production-grade content normalizer filter for Open WebUI that fixes common Markdown formatting issues in LLM outputs. It ensures that code blocks, LaTeX formulas, Mermaid diagrams, and other Markdown elements are rendered correctly.
+
+## Features
+
+*   **Mermaid Syntax Fix**: Automatically fixes common Mermaid syntax errors, such as unquoted node labels (including multi-line labels and citations) and unclosed subgraphs, ensuring diagrams render correctly.
+*   **Frontend Console Debugging**: Supports printing structured debug logs directly to the browser console (F12) for easier troubleshooting.
+*   **Code Block Formatting**: Fixes broken code block prefixes, suffixes, and indentation.
+*   **LaTeX Normalization**: Standardizes LaTeX formula delimiters (`\[` -> `$$`, `\(` -> `$`).
+*   **Thought Tag Normalization**: Unifies thought tags (`<think>`, `<thinking>` -> `<thought>`).
+*   **Escape Character Fix**: Cleans up excessive escape characters (`\\n`, `\\t`).
+*   **List Formatting**: Ensures proper newlines in list items.
+*   **Heading Fix**: Adds missing spaces in headings (`#Heading` -> `# Heading`).
+*   **Table Fix**: Adds missing closing pipes in tables.
+*   **XML Cleanup**: Removes leftover XML artifacts.
+
+## Usage
+
+1.  Install the plugin in Open WebUI.
+2.  Enable the filter globally or for specific models.
+3.  Configure the enabled fixes in the **Valves** settings.
+4.  (Optional) **Show Debug Log** is enabled by default in Valves. This prints structured logs to the browser console (F12).
+    > [!WARNING]
+    > As this is an initial version, some "negative fixes" might occur (e.g., breaking valid Markdown). If you encounter issues, please check the console logs, copy the "Original" vs "Normalized" content, and submit an issue.
+
+## Configuration (Valves)
+
+*   `priority`: Filter priority (default: 50).
+*   `enable_escape_fix`: Fix excessive escape characters.
+*   `enable_thought_tag_fix`: Normalize thought tags.
+*   `enable_code_block_fix`: Fix code block formatting.
+*   `enable_latex_fix`: Normalize LaTeX formulas.
+*   `enable_list_fix`: Fix list item newlines (Experimental).
+*   `enable_unclosed_block_fix`: Auto-close unclosed code blocks.
+*   `enable_fullwidth_symbol_fix`: Fix full-width symbols in code blocks.
+*   `enable_mermaid_fix`: Fix Mermaid syntax errors.
+*   `enable_heading_fix`: Fix missing space in headings.
+*   `enable_table_fix`: Fix missing closing pipe in tables.
+*   `enable_xml_tag_cleanup`: Cleanup leftover XML tags.
+*   `show_status`: Show status notification when fixes are applied.
+*   `show_debug_log`: Print debug logs to browser console.
+
+## License
+
+MIT

plugins/filters/markdown_normalizer/README_CN.md

@@ -0,0 +1,46 @@
+# Markdown 格式化过滤器 (Markdown Normalizer)
+
+这是一个用于 Open WebUI 的生产级内容格式化过滤器，旨在修复 LLM 输出中常见的 Markdown 格式问题。它能确保代码块、LaTeX 公式、Mermaid 图表和其他 Markdown 元素被正确渲染。
+
+## 功能特性
+
+*   **Mermaid 语法修复**: 自动修复常见的 Mermaid 语法错误，如未加引号的节点标签（支持多行标签和引用标记）和未闭合的子图 (Subgraph)，确保图表能正确渲染。
+*   **前端控制台调试**: 支持将结构化的调试日志直接打印到浏览器控制台 (F12)，方便排查问题。
+*   **代码块格式化**: 修复破损的代码块前缀、后缀和缩进问题。
+*   **LaTeX 规范化**: 标准化 LaTeX 公式定界符 (`\[` -> `$$`, `\(` -> `$`)。
+*   **思维标签规范化**: 统一思维链标签 (`<think>`, `<thinking>` -> `<thought>`)。
+*   **转义字符修复**: 清理过度的转义字符 (`\\n`, `\\t`)。
+*   **列表格式化**: 确保列表项有正确的换行。
+*   **标题修复**: 修复标题中缺失的空格 (`#标题` -> `# 标题`)。
+*   **表格修复**: 修复表格中缺失的闭合管道符。
+*   **XML 清理**: 移除残留的 XML 标签。
+
+## 使用方法
+
+1.  在 Open WebUI 中安装此插件。
+2.  全局启用或为特定模型启用此过滤器。
+3.  在 **Valves** 设置中配置需要启用的修复项。
+4.  (可选) **显示调试日志 (Show Debug Log)** 在 Valves 中默认开启。这会将结构化的日志打印到浏览器控制台 (F12)。
+    > [!WARNING]
+    > 由于这是初版，可能会出现“负向修复”的情况（例如破坏了原本正确的格式）。如果您遇到问题，请务必查看控制台日志，复制“原始 (Original)”与“规范化 (Normalized)”的内容对比，并提交 Issue 反馈。
+
+## 配置项 (Valves)
+
+*   `priority`: 过滤器优先级 (默认: 50)。
+*   `enable_escape_fix`: 修复过度的转义字符。
+*   `enable_thought_tag_fix`: 规范化思维标签。
+*   `enable_code_block_fix`: 修复代码块格式。
+*   `enable_latex_fix`: 规范化 LaTeX 公式。
+*   `enable_list_fix`: 修复列表项换行 (实验性)。
+*   `enable_unclosed_block_fix`: 自动闭合未闭合的代码块。
+*   `enable_fullwidth_symbol_fix`: 修复代码块中的全角符号。
+*   `enable_mermaid_fix`: 修复 Mermaid 语法错误。
+*   `enable_heading_fix`: 修复标题中缺失的空格。
+*   `enable_table_fix`: 修复表格中缺失的闭合管道符。
+*   `enable_xml_tag_cleanup`: 清理残留的 XML 标签。
+*   `show_status`: 应用修复时显示状态通知。
+*   `show_debug_log`: 在浏览器控制台打印调试日志。
+
+## 许可证
+
+MIT

plugins/filters/markdown_normalizer/markdown_normalizer.py

@@ -0,0 +1,531 @@
+"""
+title: Markdown Normalizer
+author: Fu-Jie
+author_url: https://github.com/Fu-Jie
+funding_url: https://github.com/Fu-Jie/awesome-openwebui
+version: 1.0.1
+description: A production-grade content normalizer filter that fixes common Markdown formatting issues in LLM outputs, such as broken code blocks, LaTeX formulas, and list formatting.
+"""
+
+from pydantic import BaseModel, Field
+from typing import Optional, List, Callable
+import re
+import logging
+import logging
+import asyncio
+import json
+from dataclasses import dataclass, field
+
+# Configure logging
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class NormalizerConfig:
+    """Configuration class for enabling/disabling specific normalization rules"""
+
+    enable_escape_fix: bool = True  # Fix excessive escape characters
+    enable_thought_tag_fix: bool = True  # Normalize thought tags
+    enable_code_block_fix: bool = True  # Fix code block formatting
+    enable_latex_fix: bool = True  # Fix LaTeX formula formatting
+    enable_list_fix: bool = (
+        False  # Fix list item newlines (default off as it can be aggressive)
+    )
+    enable_unclosed_block_fix: bool = True  # Auto-close unclosed code blocks
+    enable_fullwidth_symbol_fix: bool = False  # Fix full-width symbols in code blocks
+    enable_mermaid_fix: bool = True  # Fix common Mermaid syntax errors
+    enable_heading_fix: bool = (
+        True  # Fix missing space in headings (#Header -> # Header)
+    )
+    enable_table_fix: bool = True  # Fix missing closing pipe in tables
+    enable_xml_tag_cleanup: bool = True  # Cleanup leftover XML tags
+
+    # Custom cleaner functions (for advanced extension)
+    custom_cleaners: List[Callable[[str], str]] = field(default_factory=list)
+
+
+class ContentNormalizer:
+    """LLM Output Content Normalizer - Production Grade Implementation"""
+
+    # --- 1. Pre-compiled Regex Patterns (Performance Optimization) ---
+    _PATTERNS = {
+        # Code block prefix: if ``` is not at start of line or file
+        "code_block_prefix": re.compile(r"(?<!^)(?<!\n)(```)", re.MULTILINE),
+        # Code block suffix: ```lang followed by non-whitespace (no newline)
+        "code_block_suffix": re.compile(r"(```[\w\+\-\.]*)[ \t]+([^\n\r])"),
+        # Code block indent: whitespace at start of line + ```
+        "code_block_indent": re.compile(r"^[ \t]+(```)", re.MULTILINE),
+        # Thought tag: </thought> followed by optional whitespace/newlines
+        "thought_end": re.compile(
+            r"</(thought|think|thinking)>[ \t]*\n*", re.IGNORECASE
+        ),
+        "thought_start": re.compile(r"<(thought|think|thinking)>", re.IGNORECASE),
+        # LaTeX block: \[ ... \]
+        "latex_bracket_block": re.compile(r"\\\[(.+?)\\\]", re.DOTALL),
+        # LaTeX inline: \( ... \)
+        "latex_paren_inline": re.compile(r"\\\((.+?)\\\)"),
+        # List item: non-newline + digit + dot + space
+        "list_item": re.compile(r"([^\n])(\d+\. )"),
+        # XML artifacts (e.g. Claude's)
+        "xml_artifacts": re.compile(
+            r"</?(?:antArtifact|antThinking|artifact)[^>]*>", re.IGNORECASE
+        ),
+        # Mermaid: Match various node shapes and quote unquoted labels
+        # Fix "reverse optimization": Must precisely match shape delimiters to avoid breaking structure
+        # Priority: Longer delimiters match first
+        "mermaid_node": re.compile(
+            r'("[^"\\]*(?:\\.[^"\\]*)*")|'  # Match quoted strings first (Group 1)
+            r"(\w+)\s*(?:"
+            r"(\(\(\()(?![\"])(.*?)(?<![\"])(\)\)\))|"  # (((...))) Double Circle
+            r"(\(\()(?![\"])(.*?)(?<![\"])(\)\))|"  # ((...)) Circle
+            r"(\(\[)(?![\"])(.*?)(?<![\"])(\]\))|"  # ([...]) Stadium
+            r"(\[\()(?![\"])(.*?)(?<![\"])(\)\])|"  # [(...)] Cylinder
+            r"(\[\[)(?![\"])(.*?)(?<![\"])(\]\])|"  # [[...]] Subroutine
+            r"(\{\{)(?![\"])(.*?)(?<![\"])(\}\})|"  # {{...}} Hexagon
+            r"(\[/)(?![\"])(.*?)(?<![\"])(/\])|"  # [/.../] Parallelogram
+            r"(\[\\)(?![\"])(.*?)(?<![\"])(\\\])|"  # [\...\] Parallelogram Alt
+            r"(\[/)(?![\"])(.*?)(?<![\"])(\\\])|"  # [/...\] Trapezoid
+            r"(\[\\)(?![\"])(.*?)(?<![\"])(/\])|"  # [\.../] Trapezoid Alt
+            r"(\()(?![\"])(.*?)(?<![\"])(\))|"  # (...) Round
+            r"(\[)(?![\"])(.*?)(?<![\"])(\])|"  # [...] Square
+            r"(\{)(?![\"])(.*?)(?<![\"])(\})|"  # {...} Rhombus
+            r"(>)(?![\"])(.*?)(?<![\"])(\])"  # >...] Asymmetric
+            r")"
+            r"(\s*\[\d+\])?",  # Capture optional citation [1]
+            re.DOTALL,
+        ),
+        # Heading: #Heading -> # Heading
+        "heading_space": re.compile(r"^(#+)([^ \n#])", re.MULTILINE),
+        # Table: | col1 | col2 -> | col1 | col2 |
+        "table_pipe": re.compile(r"^(\|.*[^|\n])$", re.MULTILINE),
+    }
+
+    def __init__(self, config: Optional[NormalizerConfig] = None):
+        self.config = config or NormalizerConfig()
+        self.applied_fixes = []
+
+    def normalize(self, content: str) -> str:
+        """Main entry point: apply all normalization rules in order"""
+        self.applied_fixes = []
+        if not content:
+            return content
+
+        original_content = content  # Keep a copy for logging
+
+        try:
+            # 1. Escape character fix (Must be first)
+            if self.config.enable_escape_fix:
+                original = content
+                content = self._fix_escape_characters(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Escape Chars")
+
+            # 2. Thought tag normalization
+            if self.config.enable_thought_tag_fix:
+                original = content
+                content = self._fix_thought_tags(content)
+                if content != original:
+                    self.applied_fixes.append("Normalize Thought Tags")
+
+            # 3. Code block formatting fix
+            if self.config.enable_code_block_fix:
+                original = content
+                content = self._fix_code_blocks(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Code Blocks")
+
+            # 4. LaTeX formula normalization
+            if self.config.enable_latex_fix:
+                original = content
+                content = self._fix_latex_formulas(content)
+                if content != original:
+                    self.applied_fixes.append("Normalize LaTeX")
+
+            # 5. List formatting fix
+            if self.config.enable_list_fix:
+                original = content
+                content = self._fix_list_formatting(content)
+                if content != original:
+                    self.applied_fixes.append("Fix List Format")
+
+            # 6. Unclosed code block fix
+            if self.config.enable_unclosed_block_fix:
+                original = content
+                content = self._fix_unclosed_code_blocks(content)
+                if content != original:
+                    self.applied_fixes.append("Close Code Blocks")
+
+            # 7. Full-width symbol fix (in code blocks only)
+            if self.config.enable_fullwidth_symbol_fix:
+                original = content
+                content = self._fix_fullwidth_symbols_in_code(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Full-width Symbols")
+
+            # 8. Mermaid syntax fix
+            if self.config.enable_mermaid_fix:
+                original = content
+                content = self._fix_mermaid_syntax(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Mermaid Syntax")
+
+            # 9. Heading fix
+            if self.config.enable_heading_fix:
+                original = content
+                content = self._fix_headings(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Headings")
+
+            # 10. Table fix
+            if self.config.enable_table_fix:
+                original = content
+                content = self._fix_tables(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Tables")
+
+            # 11. XML tag cleanup
+            if self.config.enable_xml_tag_cleanup:
+                original = content
+                content = self._cleanup_xml_tags(content)
+                if content != original:
+                    self.applied_fixes.append("Cleanup XML Tags")
+
+            # 9. Custom cleaners
+            for cleaner in self.config.custom_cleaners:
+                original = content
+                content = cleaner(content)
+                if content != original:
+                    self.applied_fixes.append("Custom Cleaner")
+
+            if self.applied_fixes:
+                logger.info(f"Markdown Normalizer Applied Fixes: {self.applied_fixes}")
+                logger.debug(
+                    f"--- Original Content ---\n{original_content}\n------------------------"
+                )
+                logger.debug(
+                    f"--- Normalized Content ---\n{content}\n--------------------------"
+                )
+
+            return content
+
+        except Exception as e:
+            # Production safeguard: return original content on error
+            logger.error(f"Content normalization failed: {e}", exc_info=True)
+            return content
+
+    def _fix_escape_characters(self, content: str) -> str:
+        """Fix excessive escape characters"""
+        content = content.replace("\\r\\n", "\n")
+        content = content.replace("\\n", "\n")
+        content = content.replace("\\t", "\t")
+        content = content.replace("\\\\", "\\")
+        return content
+
+    def _fix_thought_tags(self, content: str) -> str:
+        """Normalize thought tags: unify naming and fix spacing"""
+        # 1. Standardize start tag: <think>, <thinking> -> <thought>
+        content = self._PATTERNS["thought_start"].sub("<thought>", content)
+        # 2. Standardize end tag and ensure newlines: </think> -> </thought>\n\n
+        return self._PATTERNS["thought_end"].sub("</thought>\n\n", content)
+
+    def _fix_code_blocks(self, content: str) -> str:
+        """Fix code block formatting (prefixes, suffixes, indentation)"""
+        # Remove indentation before code blocks
+        content = self._PATTERNS["code_block_indent"].sub(r"\1", content)
+        # Ensure newline before ```
+        content = self._PATTERNS["code_block_prefix"].sub(r"\n\1", content)
+        # Ensure newline after ```lang
+        content = self._PATTERNS["code_block_suffix"].sub(r"\1\n\2", content)
+        return content
+
+    def _fix_latex_formulas(self, content: str) -> str:
+        """Normalize LaTeX formulas: \[ -> $$ (block), \( -> $ (inline)"""
+        content = self._PATTERNS["latex_bracket_block"].sub(r"$$\1$$", content)
+        content = self._PATTERNS["latex_paren_inline"].sub(r"$\1$", content)
+        return content
+
+    def _fix_list_formatting(self, content: str) -> str:
+        """Fix missing newlines in lists (e.g., 'text1. item' -> 'text\\n1. item')"""
+        return self._PATTERNS["list_item"].sub(r"\1\n\2", content)
+
+    def _fix_unclosed_code_blocks(self, content: str) -> str:
+        """Auto-close unclosed code blocks"""
+        if content.count("```") % 2 != 0:
+            content += "\n```"
+        return content
+
+    def _fix_fullwidth_symbols_in_code(self, content: str) -> str:
+        """Convert full-width symbols to half-width inside code blocks"""
+        FULLWIDTH_MAP = {
+            "，": ",",
+            "。": ".",
+            "（": "(",
+            "）": ")",
+            "【": "[",
+            "】": "]",
+            "；": ";",
+            "：": ":",
+            "？": "?",
+            "！": "!",
+            '"': '"',
+            '"': '"',
+            """: "'", """: "'",
+        }
+
+        parts = content.split("```")
+        # Code block content is at odd indices: 1, 3, 5...
+        for i in range(1, len(parts), 2):
+            for full, half in FULLWIDTH_MAP.items():
+                parts[i] = parts[i].replace(full, half)
+
+        return "```".join(parts)
+
+    def _fix_mermaid_syntax(self, content: str) -> str:
+        """Fix common Mermaid syntax errors while preserving node shapes"""
+
+        def replacer(match):
+            # Group 1 is Quoted String (if matched)
+            if match.group(1):
+                return match.group(1)
+
+            # Group 2 is ID
+            id_str = match.group(2)
+
+            # Find matching shape group
+            groups = match.groups()
+            citation = groups[-1] or ""  # Last group is citation
+
+            # Iterate over shape groups (excluding the last citation group)
+            for i in range(2, len(groups) - 1, 3):
+                if groups[i] is not None:
+                    open_char = groups[i]
+                    content = groups[i + 1]
+                    close_char = groups[i + 2]
+
+                    # Append citation to content if present
+                    if citation:
+                        content += citation
+
+                    # Escape quotes in content
+                    content = content.replace('"', '\\"')
+
+                    return f'{id_str}{open_char}"{content}"{close_char}'
+
+            return match.group(0)
+
+        parts = content.split("```")
+        for i in range(1, len(parts), 2):
+            # Check if it's a mermaid block
+            lang_line = parts[i].split("\n", 1)[0].strip().lower()
+            if "mermaid" in lang_line:
+                # Apply the comprehensive regex fix
+                parts[i] = self._PATTERNS["mermaid_node"].sub(replacer, parts[i])
+
+                # Auto-close subgraphs
+                subgraph_count = len(
+                    re.findall(r"\bsubgraph\b", parts[i], re.IGNORECASE)
+                )
+                end_count = len(re.findall(r"\bend\b", parts[i], re.IGNORECASE))
+
+                if subgraph_count > end_count:
+                    missing_ends = subgraph_count - end_count
+                    parts[i] = parts[i].rstrip() + ("\n    end" * missing_ends) + "\n"
+
+        return "```".join(parts)
+
+    def _fix_headings(self, content: str) -> str:
+        """Fix missing space in headings: #Heading -> # Heading"""
+        # We only fix if it's not inside a code block.
+        # But splitting by code block is expensive.
+        # Given headings usually don't appear inside code blocks without space in valid code (except comments),
+        # we might risk false positives in comments like `#TODO`.
+        # To be safe, let's split by code blocks.
+
+        parts = content.split("```")
+        for i in range(0, len(parts), 2):  # Even indices are markdown text
+            parts[i] = self._PATTERNS["heading_space"].sub(r"\1 \2", parts[i])
+        return "```".join(parts)
+
+    def _fix_tables(self, content: str) -> str:
+        """Fix tables missing closing pipe"""
+        parts = content.split("```")
+        for i in range(0, len(parts), 2):
+            parts[i] = self._PATTERNS["table_pipe"].sub(r"\1|", parts[i])
+        return "```".join(parts)
+
+    def _cleanup_xml_tags(self, content: str) -> str:
+        """Remove leftover XML tags"""
+        return self._PATTERNS["xml_artifacts"].sub("", content)
+
+
+class Filter:
+    class Valves(BaseModel):
+        priority: int = Field(
+            default=50,
+            description="Priority level. Higher runs later (recommended to run after other filters).",
+        )
+        enable_escape_fix: bool = Field(
+            default=True, description="Fix excessive escape characters (\\n, \\t, etc.)"
+        )
+        enable_thought_tag_fix: bool = Field(
+            default=True, description="Normalize </thought> tags"
+        )
+        enable_code_block_fix: bool = Field(
+            default=True,
+            description="Fix code block formatting (indentation, newlines)",
+        )
+        enable_latex_fix: bool = Field(
+            default=True, description="Normalize LaTeX formulas (\\[ -> $$, \\( -> $)"
+        )
+        enable_list_fix: bool = Field(
+            default=False, description="Fix list item newlines (Experimental)"
+        )
+        enable_unclosed_block_fix: bool = Field(
+            default=True, description="Auto-close unclosed code blocks"
+        )
+        enable_fullwidth_symbol_fix: bool = Field(
+            default=False, description="Fix full-width symbols in code blocks"
+        )
+        enable_mermaid_fix: bool = Field(
+            default=True,
+            description="Fix common Mermaid syntax errors (e.g. unquoted labels)",
+        )
+        enable_heading_fix: bool = Field(
+            default=True,
+            description="Fix missing space in headings (#Header -> # Header)",
+        )
+        enable_table_fix: bool = Field(
+            default=True, description="Fix missing closing pipe in tables"
+        )
+        enable_xml_tag_cleanup: bool = Field(
+            default=True, description="Cleanup leftover XML tags"
+        )
+        show_status: bool = Field(
+            default=True, description="Show status notification when fixes are applied"
+        )
+        show_debug_log: bool = Field(
+            default=True, description="Print debug logs to browser console (F12)"
+        )
+
+    def __init__(self):
+        self.valves = self.Valves()
+
+    def _contains_html(self, content: str) -> bool:
+        """Check if content contains HTML tags (to avoid breaking HTML output)"""
+        pattern = r"<\s*/?\s*(?:html|head|body|div|span|p|br|hr|ul|ol|li|table|thead|tbody|tfoot|tr|td|th|img|a|b|i|strong|em|code|pre|blockquote|h[1-6]|script|style|form|input|button|label|select|option|iframe|link|meta|title)\b"
+        return bool(re.search(pattern, content, re.IGNORECASE))
+
+    async def _emit_status(self, __event_emitter__, applied_fixes: List[str]):
+        """Emit status notification"""
+        if not self.valves.show_status or not applied_fixes:
+            return
+
+        description = "✓ Markdown Normalized"
+        if applied_fixes:
+            description += f": {', '.join(applied_fixes)}"
+
+        try:
+            await __event_emitter__(
+                {
+                    "type": "status",
+                    "data": {
+                        "description": description,
+                        "done": True,
+                    },
+                }
+            )
+        except Exception as e:
+            print(f"Error emitting status: {e}")
+
+    async def _emit_debug_log(
+        self, __event_call__, applied_fixes: List[str], original: str, normalized: str
+    ):
+        """Emit debug log to browser console via JS execution"""
+        if not self.valves.show_debug_log or not __event_call__:
+            return
+
+        try:
+            # Prepare data for JS
+            log_data = {
+                "fixes": applied_fixes,
+                "original": original,
+                "normalized": normalized,
+            }
+
+            # Construct JS code
+            js_code = f"""
+                (async function() {{
+                    console.group("🛠️ Markdown Normalizer Debug");
+                    console.log("Applied Fixes:", {json.dumps(applied_fixes, ensure_ascii=False)});
+                    console.log("Original Content:", {json.dumps(original, ensure_ascii=False)});
+                    console.log("Normalized Content:", {json.dumps(normalized, ensure_ascii=False)});
+                    console.groupEnd();
+                }})();
+            """
+
+            await __event_call__(
+                {
+                    "type": "execute",
+                    "data": {"code": js_code},
+                }
+            )
+        except Exception as e:
+            print(f"Error emitting debug log: {e}")
+
+    async def outlet(
+        self,
+        body: dict,
+        __user__: Optional[dict] = None,
+        __event_emitter__=None,
+        __event_call__=None,
+        __metadata__: Optional[dict] = None,
+    ) -> dict:
+        """
+        Process the response body to normalize Markdown content.
+        """
+        if "messages" in body and body["messages"]:
+            last = body["messages"][-1]
+            content = last.get("content", "") or ""
+
+            if last.get("role") == "assistant" and isinstance(content, str):
+                # Skip if content looks like HTML to avoid breaking it
+                if self._contains_html(content):
+                    return body
+
+                # Configure normalizer based on valves
+                config = NormalizerConfig(
+                    enable_escape_fix=self.valves.enable_escape_fix,
+                    enable_thought_tag_fix=self.valves.enable_thought_tag_fix,
+                    enable_code_block_fix=self.valves.enable_code_block_fix,
+                    enable_latex_fix=self.valves.enable_latex_fix,
+                    enable_list_fix=self.valves.enable_list_fix,
+                    enable_unclosed_block_fix=self.valves.enable_unclosed_block_fix,
+                    enable_fullwidth_symbol_fix=self.valves.enable_fullwidth_symbol_fix,
+                    enable_mermaid_fix=self.valves.enable_mermaid_fix,
+                    enable_heading_fix=self.valves.enable_heading_fix,
+                    enable_table_fix=self.valves.enable_table_fix,
+                    enable_xml_tag_cleanup=self.valves.enable_xml_tag_cleanup,
+                )
+
+                normalizer = ContentNormalizer(config)
+
+                # Execute normalization
+                new_content = normalizer.normalize(content)
+
+                # Update content if changed
+                if new_content != content:
+                    last["content"] = new_content
+
+                    # Emit status if enabled
+                    if __event_emitter__:
+                        await self._emit_status(
+                            __event_emitter__, normalizer.applied_fixes
+                        )
+                        await self._emit_debug_log(
+                            __event_call__,
+                            normalizer.applied_fixes,
+                            content,
+                            new_content,
+                        )
+
+        return body

plugins/filters/markdown_normalizer/markdown_normalizer_cn.py

@@ -0,0 +1,556 @@
+"""
+title: Markdown 格式修复器 (Markdown Normalizer)
+author: Fu-Jie
+author_url: https://github.com/Fu-Jie
+funding_url: https://github.com/Fu-Jie/awesome-openwebui
+version: 1.0.1
+description: 生产级内容规范化过滤器，修复 LLM 输出中常见的 Markdown 格式问题，如损坏的代码块、LaTeX 公式、Mermaid 图表和列表格式。
+"""
+
+from pydantic import BaseModel, Field
+from typing import Optional, List, Callable
+import re
+import logging
+import asyncio
+import json
+from dataclasses import dataclass, field
+
+# Configure logging
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class NormalizerConfig:
+    """配置类，用于启用/禁用特定的规范化规则"""
+
+    enable_escape_fix: bool = True  # 修复过度的转义字符
+    enable_thought_tag_fix: bool = True  # 规范化思维链标签
+    enable_code_block_fix: bool = True  # 修复代码块格式
+    enable_latex_fix: bool = True  # 修复 LaTeX 公式格式
+    enable_list_fix: bool = False  # 修复列表项换行 (默认关闭，因为可能过于激进)
+    enable_unclosed_block_fix: bool = True  # 自动闭合未闭合的代码块
+    enable_fullwidth_symbol_fix: bool = False  # 修复代码块中的全角符号
+    enable_mermaid_fix: bool = True  # 修复常见的 Mermaid 语法错误
+    enable_heading_fix: bool = True  # 修复标题中缺失的空格 (#Header -> # Header)
+    enable_table_fix: bool = True  # 修复表格中缺失的闭合管道符
+    enable_xml_tag_cleanup: bool = True  # 清理残留的 XML 标签
+
+    # 自定义清理函数 (用于高级扩展)
+    custom_cleaners: List[Callable[[str], str]] = field(default_factory=list)
+
+
+class ContentNormalizer:
+    """LLM Output Content Normalizer - Production Grade Implementation"""
+
+    # --- 1. Pre-compiled Regex Patterns (Performance Optimization) ---
+    _PATTERNS = {
+        # Code block prefix: if ``` is not at start of line or file
+        "code_block_prefix": re.compile(r"(?<!^)(?<!\n)(```)", re.MULTILINE),
+        # Code block suffix: ```lang followed by non-whitespace (no newline)
+        "code_block_suffix": re.compile(r"(```[\w\+\-\.]*)[ \t]+([^\n\r])"),
+        # Code block indent: whitespace at start of line + ```
+        "code_block_indent": re.compile(r"^[ \t]+(```)", re.MULTILINE),
+        # Thought tag: </thought> followed by optional whitespace/newlines
+        "thought_end": re.compile(
+            r"</(thought|think|thinking)>[ \t]*\n*", re.IGNORECASE
+        ),
+        "thought_start": re.compile(r"<(thought|think|thinking)>", re.IGNORECASE),
+        # LaTeX block: \[ ... \]
+        "latex_bracket_block": re.compile(r"\\\[(.+?)\\\]", re.DOTALL),
+        # LaTeX inline: \( ... \)
+        "latex_paren_inline": re.compile(r"\\\((.+?)\\\)"),
+        # List item: non-newline + digit + dot + space
+        "list_item": re.compile(r"([^\n])(\d+\. )"),
+        # XML artifacts (e.g. Claude's)
+        "xml_artifacts": re.compile(
+            r"</?(?:antArtifact|antThinking|artifact)[^>]*>", re.IGNORECASE
+        ),
+        # Mermaid: 匹配各种形状的节点并为未加引号的标签添加引号
+        # 修复"反向优化"问题：必须精确匹配各种形状的定界符，避免破坏形状结构
+        # 优先级：长定界符优先匹配
+        "mermaid_node": re.compile(
+            r'("[^"\\]*(?:\\.[^"\\]*)*")|'  # Match quoted strings first (Group 1)
+            r"(\w+)\s*(?:"
+            r"(\(\(\()(?![\"])(.*?)(?<![\"])(\)\)\))|"  # (((...))) Double Circle
+            r"(\(\()(?![\"])(.*?)(?<![\"])(\)\))|"  # ((...)) Circle
+            r"(\(\[)(?![\"])(.*?)(?<![\"])(\]\))|"  # ([...]) Stadium
+            r"(\[\()(?![\"])(.*?)(?<![\"])(\)\])|"  # [(...)] Cylinder
+            r"(\[\[)(?![\"])(.*?)(?<![\"])(\]\])|"  # [[...]] Subroutine
+            r"(\{\{)(?![\"])(.*?)(?<![\"])(\}\})|"  # {{...}} Hexagon
+            r"(\[/)(?![\"])(.*?)(?<![\"])(/\])|"  # [/.../] Parallelogram
+            r"(\[\\)(?![\"])(.*?)(?<![\"])(\\\])|"  # [\...\] Parallelogram Alt
+            r"(\[/)(?![\"])(.*?)(?<![\"])(\\\])|"  # [/...\] Trapezoid
+            r"(\[\\)(?![\"])(.*?)(?<![\"])(/\])|"  # [\.../] Trapezoid Alt
+            r"(\()(?![\"])(.*?)(?<![\"])(\))|"  # (...) Round
+            r"(\[)(?![\"])(.*?)(?<![\"])(\])|"  # [...] Square
+            r"(\{)(?![\"])(.*?)(?<![\"])(\})|"  # {...} Rhombus
+            r"(>)(?![\"])(.*?)(?<![\"])(\])"  # >...] Asymmetric
+            r")"
+            r"(\s*\[\d+\])?",  # Capture optional citation [1]
+            re.DOTALL,
+        ),
+        # Heading: #Heading -> # Heading
+        "heading_space": re.compile(r"^(#+)([^ \n#])", re.MULTILINE),
+        # Table: | col1 | col2 -> | col1 | col2 |
+        "table_pipe": re.compile(r"^(\|.*[^|\n])$", re.MULTILINE),
+    }
+
+    def __init__(self, config: Optional[NormalizerConfig] = None):
+        self.config = config or NormalizerConfig()
+        self.applied_fixes = []
+
+    def normalize(self, content: str) -> str:
+        """Main entry point: apply all normalization rules in order"""
+        self.applied_fixes = []
+        if not content:
+            return content
+
+        original_content = content  # Keep a copy for logging
+
+        try:
+            # 1. Escape character fix (Must be first)
+            if self.config.enable_escape_fix:
+                original = content
+                content = self._fix_escape_characters(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Escape Chars")
+
+            # 2. Thought tag normalization
+            if self.config.enable_thought_tag_fix:
+                original = content
+                content = self._fix_thought_tags(content)
+                if content != original:
+                    self.applied_fixes.append("Normalize Thought Tags")
+
+            # 3. Code block formatting fix
+            if self.config.enable_code_block_fix:
+                original = content
+                content = self._fix_code_blocks(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Code Blocks")
+
+            # 4. LaTeX formula normalization
+            if self.config.enable_latex_fix:
+                original = content
+                content = self._fix_latex_formulas(content)
+                if content != original:
+                    self.applied_fixes.append("Normalize LaTeX")
+
+            # 5. List formatting fix
+            if self.config.enable_list_fix:
+                original = content
+                content = self._fix_list_formatting(content)
+                if content != original:
+                    self.applied_fixes.append("Fix List Format")
+
+            # 6. Unclosed code block fix
+            if self.config.enable_unclosed_block_fix:
+                original = content
+                content = self._fix_unclosed_code_blocks(content)
+                if content != original:
+                    self.applied_fixes.append("Close Code Blocks")
+
+            # 7. Full-width symbol fix (in code blocks only)
+            if self.config.enable_fullwidth_symbol_fix:
+                original = content
+                content = self._fix_fullwidth_symbols_in_code(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Full-width Symbols")
+
+            # 8. Mermaid syntax fix
+            if self.config.enable_mermaid_fix:
+                original = content
+                content = self._fix_mermaid_syntax(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Mermaid Syntax")
+
+            # 9. Heading fix
+            if self.config.enable_heading_fix:
+                original = content
+                content = self._fix_headings(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Headings")
+
+            # 10. Table fix
+            if self.config.enable_table_fix:
+                original = content
+                content = self._fix_tables(content)
+                if content != original:
+                    self.applied_fixes.append("Fix Tables")
+
+            # 11. XML tag cleanup
+            if self.config.enable_xml_tag_cleanup:
+                original = content
+                content = self._cleanup_xml_tags(content)
+                if content != original:
+                    self.applied_fixes.append("Cleanup XML Tags")
+
+            # 9. Custom cleaners
+            for cleaner in self.config.custom_cleaners:
+                original = content
+                content = cleaner(content)
+                if content != original:
+                    self.applied_fixes.append("Custom Cleaner")
+
+            if self.applied_fixes:
+                print(f"[Markdown Normalizer] Applied fixes: {self.applied_fixes}")
+                print(
+                    f"[Markdown Normalizer] --- Original Content ---\n{original_content}\n------------------------"
+                )
+                print(
+                    f"[Markdown Normalizer] --- Normalized Content ---\n{content}\n--------------------------"
+                )
+
+            return content
+
+        except Exception as e:
+            # Production safeguard: return original content on error
+            logger.error(f"Content normalization failed: {e}", exc_info=True)
+            return content
+
+    def _fix_escape_characters(self, content: str) -> str:
+        """Fix excessive escape characters"""
+        content = content.replace("\\r\\n", "\n")
+        content = content.replace("\\n", "\n")
+        content = content.replace("\\t", "\t")
+        content = content.replace("\\\\", "\\")
+        return content
+
+    def _fix_thought_tags(self, content: str) -> str:
+        """Normalize thought tags: unify naming and fix spacing"""
+        # 1. Standardize start tag: <think>, <thinking> -> <thought>
+        content = self._PATTERNS["thought_start"].sub("<thought>", content)
+        # 2. Standardize end tag and ensure newlines: </think> -> </thought>\n\n
+        return self._PATTERNS["thought_end"].sub("</thought>\n\n", content)
+
+    def _fix_code_blocks(self, content: str) -> str:
+        """Fix code block formatting (prefixes, suffixes, indentation)"""
+        # Remove indentation before code blocks
+        content = self._PATTERNS["code_block_indent"].sub(r"\1", content)
+        # Ensure newline before ```
+        content = self._PATTERNS["code_block_prefix"].sub(r"\n\1", content)
+        # Ensure newline after ```lang
+        content = self._PATTERNS["code_block_suffix"].sub(r"\1\n\2", content)
+        return content
+
+    def _fix_latex_formulas(self, content: str) -> str:
+        """Normalize LaTeX formulas: \[ -> $$ (block), \( -> $ (inline)"""
+        content = self._PATTERNS["latex_bracket_block"].sub(r"$$\1$$", content)
+        content = self._PATTERNS["latex_paren_inline"].sub(r"$\1$", content)
+        return content
+
+    def _fix_list_formatting(self, content: str) -> str:
+        """Fix missing newlines in lists (e.g., 'text1. item' -> 'text\\n1. item')"""
+        return self._PATTERNS["list_item"].sub(r"\1\n\2", content)
+
+    def _fix_unclosed_code_blocks(self, content: str) -> str:
+        """Auto-close unclosed code blocks"""
+        if content.count("```") % 2 != 0:
+            content += "\n```"
+        return content
+
+    def _fix_fullwidth_symbols_in_code(self, content: str) -> str:
+        """Convert full-width symbols to half-width inside code blocks"""
+        FULLWIDTH_MAP = {
+            "，": ",",
+            "。": ".",
+            "（": "(",
+            "）": ")",
+            "【": "[",
+            "】": "]",
+            "；": ";",
+            "：": ":",
+            "？": "?",
+            "！": "!",
+            '"': '"',
+            '"': '"',
+            """: "'", """: "'",
+        }
+
+        parts = content.split("```")
+        # Code block content is at odd indices: 1, 3, 5...
+        for i in range(1, len(parts), 2):
+            for full, half in FULLWIDTH_MAP.items():
+                parts[i] = parts[i].replace(full, half)
+
+        return "```".join(parts)
+
+    def _fix_mermaid_syntax(self, content: str) -> str:
+        """修复常见的 Mermaid 语法错误，同时保留节点形状"""
+
+        def replacer(match):
+            # Group 1 is Quoted String (if matched)
+            if match.group(1):
+                return match.group(1)
+
+            # Group 2 is ID
+            id_str = match.group(2)
+
+            # Find matching shape group
+            groups = match.groups()
+            citation = groups[-1] or ""  # Last group is citation
+
+            # Iterate over shape groups (excluding the last citation group)
+            for i in range(2, len(groups) - 1, 3):
+                if groups[i] is not None:
+                    open_char = groups[i]
+                    content = groups[i + 1]
+                    close_char = groups[i + 2]
+
+                    # Append citation to content if present
+                    if citation:
+                        content += citation
+
+                    # 如果内容包含引号，进行转义
+                    content = content.replace('"', '\\"')
+
+                    return f'{id_str}{open_char}"{content}"{close_char}'
+
+            return match.group(0)
+
+        parts = content.split("```")
+        for i in range(1, len(parts), 2):
+            # Check if it's a mermaid block
+            lang_line = parts[i].split("\n", 1)[0].strip().lower()
+            if "mermaid" in lang_line:
+                # Apply the comprehensive regex fix
+                parts[i] = self._PATTERNS["mermaid_node"].sub(replacer, parts[i])
+
+                # Auto-close subgraphs
+                # Count 'subgraph' and 'end' (case-insensitive)
+                # We use a simple regex to avoid matching words inside labels (though labels are now quoted, so it's safer)
+                # But for simplicity and speed, we just count occurrences in the whole block.
+                # A more robust way would be to strip quoted strings first, but that's expensive.
+                # Given we just quoted everything, let's try to count keywords outside quotes?
+                # Actually, since we just normalized nodes, most text is in quotes.
+                # Let's just do a simple count. It's a heuristic fix.
+                subgraph_count = len(
+                    re.findall(r"\bsubgraph\b", parts[i], re.IGNORECASE)
+                )
+                end_count = len(re.findall(r"\bend\b", parts[i], re.IGNORECASE))
+
+                if subgraph_count > end_count:
+                    missing_ends = subgraph_count - end_count
+                    parts[i] = parts[i].rstrip() + ("\n    end" * missing_ends) + "\n"
+
+        return "```".join(parts)
+
+    def _fix_headings(self, content: str) -> str:
+        """Fix missing space in headings: #Heading -> # Heading"""
+        # We only fix if it's not inside a code block.
+        # But splitting by code block is expensive.
+        # Given headings usually don't appear inside code blocks without space in valid code (except comments),
+        # we might risk false positives in comments like `#TODO`.
+        # To be safe, let's split by code blocks.
+
+        parts = content.split("```")
+        for i in range(0, len(parts), 2):  # Even indices are markdown text
+            parts[i] = self._PATTERNS["heading_space"].sub(r"\1 \2", parts[i])
+        return "```".join(parts)
+
+    def _fix_tables(self, content: str) -> str:
+        """Fix tables missing closing pipe"""
+        parts = content.split("```")
+        for i in range(0, len(parts), 2):
+            parts[i] = self._PATTERNS["table_pipe"].sub(r"\1|", parts[i])
+        return "```".join(parts)
+
+    def _cleanup_xml_tags(self, content: str) -> str:
+        """Remove leftover XML tags"""
+        return self._PATTERNS["xml_artifacts"].sub("", content)
+
+
+class Filter:
+    class Valves(BaseModel):
+        priority: int = Field(
+            default=50,
+            description="优先级。数值越高运行越晚 (建议在其他过滤器之后运行)。",
+        )
+        enable_escape_fix: bool = Field(
+            default=True, description="修复过度的转义字符 (\\n, \\t 等)"
+        )
+        enable_thought_tag_fix: bool = Field(
+            default=True, description="规范化思维链标签 (<think> -> <thought>)"
+        )
+        enable_code_block_fix: bool = Field(
+            default=True,
+            description="修复代码块格式 (缩进、换行)",
+        )
+        enable_latex_fix: bool = Field(
+            default=True, description="规范化 LaTeX 公式 (\\[ -> $$, \\( -> $)"
+        )
+        enable_list_fix: bool = Field(
+            default=False, description="修复列表项换行 (实验性)"
+        )
+        enable_unclosed_block_fix: bool = Field(
+            default=True, description="自动闭合未闭合的代码块"
+        )
+        enable_fullwidth_symbol_fix: bool = Field(
+            default=False, description="修复代码块中的全角符号"
+        )
+        enable_mermaid_fix: bool = Field(
+            default=True,
+            description="修复常见的 Mermaid 语法错误 (如未加引号的标签)",
+        )
+        enable_heading_fix: bool = Field(
+            default=True,
+            description="修复标题中缺失的空格 (#Header -> # Header)",
+        )
+        enable_table_fix: bool = Field(
+            default=True, description="修复表格中缺失的闭合管道符"
+        )
+        enable_xml_tag_cleanup: bool = Field(
+            default=True, description="清理残留的 XML 标签"
+        )
+        show_status: bool = Field(default=True, description="应用修复时显示状态通知")
+        show_debug_log: bool = Field(
+            default=True, description="在浏览器控制台打印调试日志 (F12)"
+        )
+
+    def __init__(self):
+        self.valves = self.Valves()
+
+    def _contains_html(self, content: str) -> bool:
+        """Check if content contains HTML tags (to avoid breaking HTML output)"""
+        pattern = r"<\s*/?\s*(?:html|head|body|div|span|p|br|hr|ul|ol|li|table|thead|tbody|tfoot|tr|td|th|img|a|b|i|strong|em|code|pre|blockquote|h[1-6]|script|style|form|input|button|label|select|option|iframe|link|meta|title)\b"
+        return bool(re.search(pattern, content, re.IGNORECASE))
+
+    async def _emit_status(self, __event_emitter__, applied_fixes: List[str]):
+        """Emit status notification"""
+        if not self.valves.show_status or not applied_fixes:
+            return
+
+        description = "✓ Markdown 已修复"
+        if applied_fixes:
+            # Translate fix names for status display
+            fix_map = {
+                "Fix Escape Chars": "转义字符",
+                "Normalize Thought Tags": "思维标签",
+                "Fix Code Blocks": "代码块",
+                "Normalize LaTeX": "LaTeX公式",
+                "Fix List Format": "列表格式",
+                "Close Code Blocks": "闭合代码块",
+                "Fix Full-width Symbols": "全角符号",
+                "Fix Mermaid Syntax": "Mermaid语法",
+                "Fix Headings": "标题格式",
+                "Fix Tables": "表格格式",
+                "Cleanup XML Tags": "XML清理",
+                "Custom Cleaner": "自定义清理",
+            }
+            translated_fixes = [fix_map.get(fix, fix) for fix in applied_fixes]
+            description += f": {', '.join(translated_fixes)}"
+
+        try:
+            await __event_emitter__(
+                {
+                    "type": "status",
+                    "data": {
+                        "description": description,
+                        "done": True,
+                    },
+                }
+            )
+        except Exception as e:
+            print(f"Error emitting status: {e}")
+
+    async def _emit_debug_log(
+        self,
+        __event_emitter__,
+        applied_fixes: List[str],
+        original: str,
+        normalized: str,
+    ):
+        """Emit debug log to browser console via JS execution"""
+
+    async def _emit_debug_log(
+        self, __event_call__, applied_fixes: List[str], original: str, normalized: str
+    ):
+        """Emit debug log to browser console via JS execution"""
+        if not self.valves.show_debug_log or not __event_call__:
+            return
+
+        try:
+            # Prepare data for JS
+            log_data = {
+                "fixes": applied_fixes,
+                "original": original,
+                "normalized": normalized,
+            }
+
+            # Construct JS code
+            js_code = f"""
+                (async function() {{
+                    console.group("🛠️ Markdown Normalizer Debug");
+                    console.log("Applied Fixes:", {json.dumps(applied_fixes, ensure_ascii=False)});
+                    console.log("Original Content:", {json.dumps(original, ensure_ascii=False)});
+                    console.log("Normalized Content:", {json.dumps(normalized, ensure_ascii=False)});
+                    console.groupEnd();
+                }})();
+            """
+            await __event_call__(
+                {
+                    "type": "execute",
+                    "data": {"code": js_code},
+                }
+            )
+
+        except Exception as e:
+            print(f"Error emitting debug log: {e}")
+
+    async def outlet(
+        self,
+        body: dict,
+        __user__: Optional[dict] = None,
+        __event_emitter__=None,
+        __event_call__=None,
+        __metadata__: Optional[dict] = None,
+    ) -> dict:
+        """
+        Process the response body to normalize Markdown content.
+        """
+        if "messages" in body and body["messages"]:
+            last = body["messages"][-1]
+            content = last.get("content", "") or ""
+
+            if last.get("role") == "assistant" and isinstance(content, str):
+                # Skip if content looks like HTML to avoid breaking it
+                if self._contains_html(content):
+                    return body
+
+                # Configure normalizer based on valves
+                config = NormalizerConfig(
+                    enable_escape_fix=self.valves.enable_escape_fix,
+                    enable_thought_tag_fix=self.valves.enable_thought_tag_fix,
+                    enable_code_block_fix=self.valves.enable_code_block_fix,
+                    enable_latex_fix=self.valves.enable_latex_fix,
+                    enable_list_fix=self.valves.enable_list_fix,
+                    enable_unclosed_block_fix=self.valves.enable_unclosed_block_fix,
+                    enable_fullwidth_symbol_fix=self.valves.enable_fullwidth_symbol_fix,
+                    enable_mermaid_fix=self.valves.enable_mermaid_fix,
+                    enable_heading_fix=self.valves.enable_heading_fix,
+                    enable_table_fix=self.valves.enable_table_fix,
+                    enable_xml_tag_cleanup=self.valves.enable_xml_tag_cleanup,
+                )
+
+                normalizer = ContentNormalizer(config)
+
+                # Execute normalization
+                new_content = normalizer.normalize(content)
+
+                # Update content if changed
+                if new_content != content:
+                    last["content"] = new_content
+
+                    # Emit status if enabled
+                    if __event_emitter__:
+                        await self._emit_status(
+                            __event_emitter__, normalizer.applied_fixes
+                        )
+                        await self._emit_debug_log(
+                            __event_call__,
+                            normalizer.applied_fixes,
+                            content,
+                            new_content,
+                        )
+
+        return body

plugins/filters/markdown_normalizer/test_markdown_normalizer.py

@@ -0,0 +1,191 @@
+import unittest
+import sys
+import os
+
+# Add the current directory to sys.path to import the module
+current_dir = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(current_dir)
+
+from markdown_normalizer import ContentNormalizer, NormalizerConfig
+
+
+class TestMarkdownNormalizer(unittest.TestCase):
+    def setUp(self):
+        self.config = NormalizerConfig(
+            enable_escape_fix=True,
+            enable_thought_tag_fix=True,
+            enable_code_block_fix=True,
+            enable_latex_fix=True,
+            enable_list_fix=True,
+            enable_unclosed_block_fix=True,
+            enable_fullwidth_symbol_fix=True,
+            enable_mermaid_fix=True,
+            enable_xml_tag_cleanup=True,
+        )
+        self.normalizer = ContentNormalizer(self.config)
+
+    def test_escape_fix(self):
+        input_text = "Line 1\\nLine 2\\tTabbed"
+        expected = "Line 1\nLine 2\tTabbed"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_thought_tag_fix(self):
+        # Case 1: Standard tag spacing
+        input_text = "Thinking...</thought>Result"
+        expected = "Thinking...</thought>\n\nResult"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+        # Case 2: Tag standardization (<think> -> <thought>)
+        input_text_deepseek = "<think>Deep thinking...</think>Result"
+        expected_deepseek = "<thought>Deep thinking...</thought>\n\nResult"
+        self.assertEqual(
+            self.normalizer.normalize(input_text_deepseek), expected_deepseek
+        )
+
+    def test_code_block_fix(self):
+        # Case 1: Indentation
+        self.assertEqual(self.normalizer._fix_code_blocks("   ```python"), "```python")
+
+        # Case 2: Prefix (newline before block)
+        self.assertEqual(
+            self.normalizer._fix_code_blocks("Text```python"), "Text\n```python"
+        )
+
+        # Case 3: Suffix (newline after lang)
+        self.assertEqual(
+            self.normalizer._fix_code_blocks("```python print('hi')"),
+            "```python\nprint('hi')",
+        )
+
+    def test_latex_fix(self):
+        input_text = "Block: \\[ x^2 \\] Inline: \\( E=mc^2 \\)"
+        expected = "Block: $$ x^2 $$ Inline: $ E=mc^2 $"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_list_fix(self):
+        input_text = "Item 1. First\nItem 2. Second"  # This is fine
+        input_text_bad = "Header1. Item 1"
+        expected = "Header\n1. Item 1"
+        self.assertEqual(self.normalizer.normalize(input_text_bad), expected)
+
+    def test_unclosed_code_block_fix(self):
+        input_text = "```python\nprint('hello')"
+        expected = "```python\nprint('hello')\n```"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_fullwidth_symbol_fix(self):
+        input_text = "Outside：Fullwidth ```python\nprint（'hello'）```"
+        expected = "Outside：Fullwidth \n```python\nprint('hello')\n```"
+
+        normalized = self.normalizer.normalize(input_text)
+        self.assertIn("print('hello')", normalized)
+        self.assertIn("Outside：Fullwidth", normalized)
+        self.assertNotIn("（", normalized)
+        self.assertNotIn("）", normalized)
+
+    def test_mermaid_fix(self):
+        # Test Mermaid syntax fix for unquoted labels
+        # Note: Regex-based fix handles mixed brackets well (e.g. [] inside ())
+        # but cannot perfectly handle same-type nesting (e.g. {} inside {}) without a parser.
+        input_text = """
+```mermaid
+graph TD
+    A[Label with (parens)] --> B(Label with [brackets])
+    C{Label with [brackets]}
+```
+"""
+        expected_snippet = """
+```mermaid
+graph TD
+    A["Label with (parens)"] --> B("Label with [brackets]")
+    C{"Label with [brackets]"}
+```
+"""
+        normalized = self.normalizer.normalize(input_text)
+
+        self.assertIn('A["Label with (parens)"]', normalized)
+        self.assertIn('B("Label with [brackets]")', normalized)
+        self.assertIn('C{"Label with [brackets]"}', normalized)
+
+    def test_mermaid_shapes_regression(self):
+        # Regression test for "reverse optimization" where ((...)) was broken into ("(...)")
+        input_text = """
+```mermaid
+graph TD
+    Start((开始)) --> Input[[输入]]
+    Input --> Verify{验证}
+    Verify --> End(((结束)))
+```
+"""
+        expected_snippet = """
+```mermaid
+graph TD
+    Start(("开始")) --> Input[["输入"]]
+    Input --> Verify{"验证"}
+    Verify --> End((("结束")))
+```
+"""
+        normalized = self.normalizer.normalize(input_text)
+        self.assertIn('Start(("开始"))', normalized)
+        self.assertIn('Input[["输入"]]', normalized)
+        self.assertIn('Verify{"验证"}', normalized)
+        self.assertIn('End((("结束")))', normalized)
+
+    def test_xml_cleanup(self):
+        input_text = "Some text <antArtifact>hidden</antArtifact> visible"
+        expected = "Some text hidden visible"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_heading_fix(self):
+        input_text = "#Heading 1\n##Heading 2\n### Valid Heading"
+        expected = "# Heading 1\n## Heading 2\n### Valid Heading"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_table_fix(self):
+        input_text = "| Col 1 | Col 2\n| Val 1 | Val 2"
+        expected = "| Col 1 | Col 2|\n| Val 1 | Val 2|"
+        self.assertEqual(self.normalizer.normalize(input_text), expected)
+
+    def test_mermaid_subgraph_autoclose(self):
+        """Test auto-closing of Mermaid subgraphs"""
+        # Case 1: Simple unclosed subgraph
+        original = """
+```mermaid
+graph TD
+    subgraph One
+        A --> B
+```
+"""
+        expected = """
+```mermaid
+graph TD
+    subgraph One
+        A --> B
+    end
+```
+"""
+        # Note: The normalizer might add quotes to A and B if they match the node pattern,
+        # but here they are simple IDs. However, our regex is strict about shapes.
+        # Simple IDs like A and B are NOT matched by our mermaid_node regex because it requires a shape delimiter.
+        # So A and B remain A and B.
+
+        normalized = self.normalizer.normalize(original)
+        # We need to be careful about whitespace in comparison
+        self.assertIn("end", normalized)
+        self.assertEqual(normalized.strip(), expected.strip())
+
+        # Case 2: Nested unclosed subgraphs
+        original_nested = """
+```mermaid
+graph TD
+    subgraph Outer
+        subgraph Inner
+            C --> D
+```
+"""
+        normalized_nested = self.normalizer.normalize(original_nested)
+        self.assertEqual(normalized_nested.count("end"), 2)
+
+
+if __name__ == "__main__":
+    unittest.main()

plugins/插件开发总结.md

@@ -41,19 +41,19 @@
 
 ## Actions (动作插件)
 
-1. **📊 智能信息图 (infographic/信息图.py)** - 基于 AntV Infographic 的智能信息图生成插件,支持多种专业模板与 SVG/PNG 下载
+1. **📊 智能信息图 (infographic/infographic_cn.py)** - 基于 AntV Infographic 的智能信息图生成插件,支持多种专业模板与 SVG/PNG 下载
 
-2. **🧠 思维导图 (smart-mind-map/思维导图.py)** - 智能分析文本内容生成交互式思维导图,帮助用户结构化和可视化知识
+2. **🧠 思维导图 (smart-mind-map/smart_mind_map_cn.py)** - 智能分析文本内容生成交互式思维导图,帮助用户结构化和可视化知识
 
-3. **📊 导出为 Excel (export_to_excel/导出为Excel.py)** - 将对话历史中的 Markdown 表格导出为符合中国规范的 Excel 文件
+3. **📊 导出为 Excel (export_to_excel/export_to_excel_cn.py)** - 将对话历史中的 Markdown 表格导出为符合中国规范的 Excel 文件
 
-4. **⚡ 闪记卡 (knowledge-card/闪记卡.py)** - 快速将文本提炼为精美的学习记忆卡片,支持核心要点提取与分类
+4. **⚡ 闪记卡 (flash-card/flash_card_cn.py)** - 快速将文本提炼为精美的学习记忆卡片,支持核心要点提取与分类
 
-5. **📖 精读 (summary/精读.py)** - 深度分析长篇文本,提炼详细摘要、关键信息点和可执行的行动建议
+5. **📖 精读 (summary/summary_cn.py)** - 深度分析长篇文本,提炼详细摘要、关键信息点和可执行的行动建议
 
 ## Filters (过滤器插件)
 
-1. **🔄 异步上下文压缩 (async-context-compression/异步上下文压缩.py)** - 异步生成摘要并压缩对话历史,支持数据库持久化存储
+1. **🔄 异步上下文压缩 (async-context-compression/async_context_compression_cn.py)** - 异步生成摘要并压缩对话历史,支持数据库持久化存储
 
 2. **✨ 上下文增强过滤器 (context_enhancement_filter/context_enhancement_filter.py)** - 增强请求上下文和优化模型功能,包含环境变量管理、模型功能适配和内容清洗
 

scripts/download_plugin_images.py

@@ -0,0 +1,133 @@
+"""
+Download plugin images from OpenWebUI Community
+下载远程插件图片到本地目录
+"""
+
+import os
+import sys
+import re
+import requests
+from urllib.parse import urlparse
+
+# Add current directory to path
+sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+
+from openwebui_community_client import get_client
+
+
+def find_local_plugin_by_id(plugins_dir: str, post_id: str) -> str | None:
+    """根据 post_id 查找本地插件文件"""
+    for root, _, files in os.walk(plugins_dir):
+        for file in files:
+            if file.endswith(".py"):
+                file_path = os.path.join(root, file)
+                with open(file_path, "r", encoding="utf-8") as f:
+                    content = f.read(2000)
+
+                id_match = re.search(
+                    r"(?:openwebui_id|post_id):\s*([a-z0-9-]+)", content
+                )
+                if id_match and id_match.group(1).strip() == post_id:
+                    return file_path
+    return None
+
+
+def download_image(url: str, save_path: str) -> bool:
+    """下载图片"""
+    try:
+        response = requests.get(url, timeout=30)
+        response.raise_for_status()
+        with open(save_path, "wb") as f:
+            f.write(response.content)
+        return True
+    except Exception as e:
+        print(f"  Error downloading: {e}")
+        return False
+
+
+def get_image_extension(url: str) -> str:
+    """从 URL 获取图片扩展名"""
+    parsed = urlparse(url)
+    path = parsed.path
+    ext = os.path.splitext(path)[1].lower()
+    if ext in [".png", ".jpg", ".jpeg", ".gif", ".webp"]:
+        return ext
+    return ".png"  # 默认
+
+
+def main():
+    try:
+        client = get_client()
+    except ValueError as e:
+        print(f"Error: {e}")
+        sys.exit(1)
+
+    base_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+    plugins_dir = os.path.join(base_dir, "plugins")
+
+    print("Fetching remote posts from OpenWebUI Community...")
+    posts = client.get_all_posts()
+    print(f"Found {len(posts)} remote posts.\n")
+
+    downloaded = 0
+    skipped = 0
+    not_found = 0
+
+    for post in posts:
+        post_id = post.get("id")
+        title = post.get("title", "Unknown")
+        media = post.get("media", [])
+
+        if not media:
+            continue
+
+        # 只取第一张图片
+        first_media = media[0] if isinstance(media, list) else media
+
+        # 处理字典格式 {'url': '...', 'type': 'image'}
+        if isinstance(first_media, dict):
+            image_url = first_media.get("url")
+        else:
+            image_url = first_media
+
+        if not image_url:
+            continue
+
+        print(f"Processing: {title}")
+        print(f"  Image URL: {image_url}")
+
+        # 查找对应的本地插件
+        local_plugin = find_local_plugin_by_id(plugins_dir, post_id)
+        if not local_plugin:
+            print(f"  ⚠️  No local plugin found for ID: {post_id}")
+            not_found += 1
+            continue
+
+        # 确定保存路径
+        plugin_dir = os.path.dirname(local_plugin)
+        plugin_name = os.path.splitext(os.path.basename(local_plugin))[0]
+        ext = get_image_extension(image_url)
+        save_path = os.path.join(plugin_dir, plugin_name + ext)
+
+        # 检查是否已存在
+        if os.path.exists(save_path):
+            print(f"  ⏭️  Image already exists: {os.path.basename(save_path)}")
+            skipped += 1
+            continue
+
+        # 下载
+        print(f"  Downloading to: {save_path}")
+        if download_image(image_url, save_path):
+            print(f"  ✅ Downloaded: {os.path.basename(save_path)}")
+            downloaded += 1
+        else:
+            print(f"  ❌ Failed to download")
+
+    print(f"\n{'='*50}")
+    print(
+        f"Finished: {downloaded} downloaded, {skipped} skipped, {not_found} not found locally"
+    )
+
+
+if __name__ == "__main__":
+    main()