cso1z/Feishu-MCP

What it solves

Feishu-MCP allows AI coding tools (like Cursor, Windsurf, and Cline) to interact directly with Feishu (Lark) documents, tasks, and user information. It bridges the gap between AI agents and the Feishu ecosystem, enabling them to read, create, and edit documents and manage project tasks without manual copying and pasting.

How it works

The project implements a server based on the Model Context Protocol (MCP), which provides a standardized way for AI models to access external tools. It supports multiple connection modes including stdio, SSE, and HTTP Streamable. It can operate using either application-level (tenant) authentication or user-level (user) authentication via OAuth, allowing the AI to act on behalf of a specific user.

Who it’s for

Developers and teams using Feishu/Lark for documentation and task management who want to automate their workflow by integrating these resources into their AI-powered IDEs or agents.

Highlights

• Comprehensive Document Control: Create, edit, search, and structure Feishu documents, including support for rich text, tables, and LaTeX formulas.
• Task & User Management: Create, update, and delete tasks (including sub-tasks) and search for user information to assign responsibilities.
• Visual Content Support: Ability to insert images and create complex diagrams using Mermaid and Feishu's native whiteboard features.
• Flexible Deployment: Available as an MCP server or a standalone CLI tool (feishu-tool) for use in scripts or other AI agents.
• Lark International Support: Fully compatible with Lark international versions through simple configuration.