Feishu MCP Server

Enables AI assistants to directly read, write, and manage Feishu documents

README.md

🔗 Feishu MCP Server

让 AI 助手直接读写飞书文档

English | 中文


✨ 简介

Feishu MCP Server 是一个基于 Model Context Protocol (MCP) 的飞书文档集成服务。它让 AI 助手(如 GitHub Copilot、Claude、Cursor 等)能够直接操作飞书文档,实现文档自动化。

🎯 应用场景

  • 📝 技术文档自动生成 - AI 自动撰写设计文档、API 文档
  • 📊 数据报表输出 - 将数据分析结果直接写入飞书
  • 📋 会议纪要整理 - 自动生成结构化的会议记录
  • 🔄 知识库维护 - 批量更新和管理知识库内容

🚀 功能特性

📄 文档操作

功能 工具 说明
读取文档 feishu_read_document 支持 Markdown/JSON 格式输出
创建文档 feishu_create_document 支持完整 Markdown 语法
追加内容 feishu_append_content 智能识别 Markdown 表格
更新内容 feishu_update_content 替换文档全部内容
文档信息 feishu_get_document_info 获取标题、ID、版本等

📊 富文本元素

功能 工具 说明
插入表格 feishu_insert_table 自适应列宽,支持 Markdown 表格语法
插入图表 feishu_insert_diagram Mermaid 流程图、时序图、架构图
插入引用块 feishu_insert_callout 高亮提示信息
上传图片 feishu_upload_image 上传本地图片
插入图片 feishu_insert_image 将图片插入文档

📈 电子表格 & 多维表格

功能 工具 说明
创建电子表格 feishu_create_spreadsheet 创建新的 Sheet
写入数据 feishu_write_spreadsheet 批量写入单元格数据
读取数据 feishu_read_spreadsheet 读取指定范围数据
创建多维表格 feishu_create_bitable 创建 Bitable
添加记录 feishu_add_bitable_records 批量添加记录
读取记录 feishu_read_bitable 读取表格数据

🔧 块操作

功能 工具 说明
列出块 feishu_list_blocks 查看文档结构
删除块 feishu_delete_block 删除指定块

📦 快速开始

1. 安装

# 克隆项目
git clone https://github.com/redleaves/feishu-mcp-server.git
cd feishu-mcp-server

# 安装依赖
npm install

# 编译
npm run build

2. 配置飞书应用

  1. 访问 飞书开放平台 创建应用
  2. 获取 App IDApp Secret
  3. 添加以下权限并发布应用:
权限标识 权限名称 用途
docx:document 查看、评论和编辑新版文档 文档读写
docx:document:readonly 查看新版文档 文档读取
wiki:wiki:readonly 查看知识库 知识库访问
drive:drive 查看、评论和编辑云空间中的文件 云空间操作
bitable:bitable 查看、评论和编辑多维表格 多维表格操作

3. 配置 MCP 客户端

🐙 GitHub Copilot CLI

编辑 ~/.copilot/mcp-config.json

{
  "mcpServers": {
    "feishu": {
      "type": "stdio",
      "command": "node",
      "args": ["/path/to/feishu-mcp-server/dist/index.js"],
      "env": {
        "FEISHU_APP_ID": "cli_xxxxxxxx",
        "FEISHU_APP_SECRET": "your_app_secret"
      }
    }
  }
}
🤖 Claude Desktop

编辑 ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "feishu": {
      "command": "node",
      "args": ["/path/to/feishu-mcp-server/dist/index.js"],
      "env": {
        "FEISHU_APP_ID": "cli_xxxxxxxx",
        "FEISHU_APP_SECRET": "your_app_secret"
      }
    }
  }
}
✨ Cursor

在 Cursor 设置中添加 MCP Server:

{
  "mcp.servers": {
    "feishu": {
      "command": "node",
      "args": ["/path/to/feishu-mcp-server/dist/index.js"],
      "env": {
        "FEISHU_APP_ID": "cli_xxxxxxxx",
        "FEISHU_APP_SECRET": "your_app_secret"
      }
    }
  }
}

💡 使用示例

创建技术设计文档

创建一个飞书文档,标题是"系统架构设计",包含:
1. 项目背景
2. 技术方案(插入架构图)
3. 实施计划表格

读取并总结文档

读取这个飞书文档 https://xxx.feishu.cn/wiki/ABC123,帮我总结要点

插入数据表格

在文档中插入一个表格,展示各模块的开发进度

生成架构图

在文档中插入一个 Mermaid 流程图,展示用户登录流程

📐 技术架构

flowchart TB
    subgraph Clients["AI 客户端"]
        Copilot["GitHub Copilot"]
        Claude["Claude Desktop"]
        Cursor["Cursor"]
    end
    
    subgraph MCP["Feishu MCP Server"]
        Tools["18 个工具"]
        Auth["Token 管理"]
        API["API 封装"]
    end
    
    subgraph Feishu["飞书开放平台"]
        Doc["文档 API"]
        Drive["云空间 API"]
        Bitable["多维表格 API"]
    end
    
    Clients --> MCP
    MCP --> Feishu

⚠️ 已知限制

限制项 说明
API 频率 每秒最多 3 次请求
表格行数 单次创建最多 9 行(超出自动截断)
文件上传 最大 20MB
Token 有效期 2 小时(自动刷新)

🛠️ 开发指南

# 开发模式(热重载)
npm run dev

# 编译
npm run build

# 运行测试
npm test

# 代码检查
npm run lint

项目结构

feishu-mcp-server/
├── src/
│   ├── index.ts          # 入口文件
│   ├── constants.ts      # 常量定义
│   ├── types.ts          # 类型定义
│   ├── services/         # 服务层
│   │   ├── api.ts        # API 封装
│   │   └── auth.ts       # 认证服务
│   └── tools/            # MCP 工具
│       ├── document.ts   # 文档操作
│       ├── table.ts      # 表格操作
│       ├── block.ts      # 块操作
│       ├

Tools 18

feishu_read_documentReads a Feishu document and outputs in Markdown or JSON format.
feishu_create_documentCreates a new Feishu document using Markdown syntax.
feishu_append_contentAppends content to an existing Feishu document.
feishu_update_contentReplaces the entire content of a Feishu document.
feishu_get_document_infoRetrieves metadata including title, ID, and version.
feishu_insert_tableInserts a table into a document with support for Markdown syntax.
feishu_insert_diagramInserts a Mermaid diagram into a document.
feishu_insert_calloutInserts a highlighted callout block into a document.
feishu_upload_imageUploads a local image to Feishu.
feishu_insert_imageInserts an uploaded image into a document.
feishu_create_spreadsheetCreates a new Feishu spreadsheet.
feishu_write_spreadsheetBatch writes data to spreadsheet cells.
feishu_read_spreadsheetReads data from a specified spreadsheet range.
feishu_create_bitableCreates a new Bitable.
feishu_add_bitable_recordsBatch adds records to a Bitable.
feishu_read_bitableReads data from a Bitable.
feishu_list_blocksLists blocks within a document to view structure.
feishu_delete_blockDeletes a specific block from a document.

Environment Variables

FEISHU_APP_IDrequiredThe App ID from the Feishu Open Platform
FEISHU_APP_SECRETrequiredThe App Secret from the Feishu Open Platform

Try it

Create a new Feishu document titled 'Project Roadmap' and add a section for Q3 goals.
Read the Feishu document at [URL] and summarize the key action items.
Insert a Mermaid sequence diagram into my document showing the user authentication flow.
Add a new row to my 'Task Tracker' Bitable with the status set to 'In Progress'.
Create a spreadsheet to track monthly expenses and populate it with the provided data.

Frequently Asked Questions

What are the key features of Feishu MCP Server?

Direct read and write access to Feishu documents using Markdown.. Support for rich text elements including Mermaid diagrams, tables, and callouts.. Full management of Feishu spreadsheets and Bitable (multi-dimensional tables).. Granular document structure control via block operations.. Automated image uploading and insertion into documents..

What can I use Feishu MCP Server for?

Automating the generation of technical design and API documentation.. Exporting AI-processed data analysis results directly into Feishu spreadsheets.. Automatically generating structured meeting minutes from raw notes.. Batch updating and maintaining internal knowledge base content..

How do I install Feishu MCP Server?

Install Feishu MCP Server by running: git clone https://github.com/redleaves/feishu-mcp-server.git && cd feishu-mcp-server && npm install && npm run build

What MCP clients work with Feishu MCP Server?

Feishu MCP Server works with any MCP-compatible client including Claude Desktop, Claude Code, Cursor, and other editors with MCP support.

Turn this server into reusable context

Keep Feishu MCP Server docs, env vars, and workflow notes in Conare so your agent carries them across sessions.

Open Conare