lanhu-mcp
⚡ 需求分析效率提升 200%!全球首个为 AI 编程时代设计的团队协作 MCP 服务器,自动分析需求自动编写前后端代码,下载切图
What is lanhu-mcp?
⚡ 需求分析效率提升 200%!全球首个为 AI 编程时代设计的团队协作 MCP 服务器,自动分析需求自动编写前后端代码,下载切图
Overview of lanhu-mcp
What it is
Lanhu MCP Server is a Model Context Protocol (MCP) server designed for Lanhu (蓝湖) design delivery and requirement reading. It allows AI assistants to share team knowledge by connecting to a central knowledge hub, breaking the isolation of individual AI IDEs. The server provides source data and resources via MCP, enabling large models to understand design visuals and adapt to target engineering projects.
Key features
- Requirement document analysis: Extracts Axure pages, text, and annotations; offers three analysis modes (development, testing, quick exploration); uses a four-stage workflow (global scan, group analysis, reverse verification, deliverable generation); supports long-page visual evidence with full screenshots and overlapping high-resolution chunks.
- UI design support: Reads designs by screen and layer ID; delivers original sliced images with size/hash verification; batch downloads and displays UI designs; returns design analysis with preview, source parameters, and schema-derived HTML+CSS reference; automatically identifies and exports design slices and icon resources; maps layers and assets by stable IDs.
- Team collaboration message board: Unifies knowledge base for all AI assistants; enables context passing between different AI tools; allows permanent storage of pitfalls, experiences, and best practices as knowledge base entries; supports task-type messages for querying code or databases; integrates Feishu (飞书) notifications for @mentions; tracks which AI accessed which documents.
- Performance optimization: Version-based smart caching for permanent cache; incremental updates to download only changed resources; concurrent processing for batch page screenshots and resource downloads.
Use cases
- Integrating with AI development tools that support the MCP protocol, such as Cursor, Windsurf, Claude Code, OpenClaw, ClawBot, Trae, 通义灵码 (Tongyi Lingma), and Cline, to directly read Lanhu requirement documents and design drafts.
- Enabling team-wide AI collaboration: one developer's AI analyzes requirements, and other team members' AIs can reuse the analysis; knowledge is shared across the team instead of being isolated in individual IDEs.
Getting started
Two installation methods are available: Method 1: Let AI install (recommended) Tell your AI: Help me clone and install https://github.com/dsphper/lanhu-mcp Method 2: Manual installation Docker deployment (recommended): bash git clone https://github.com/dsphper/lanhu-mcp.git cd lanhu-mcp cp .env.example .env # Edit .env and set LANHU_COOKIE docker compose up -d --build Source code run (Python 3.10+ required): bash git clone https://github.com/dsphper/lanhu-mcp.git cd lanhu-mcp bash easy-install.sh # Linux/Mac # or easy-install.bat on Windows After installation, configure the Lanhu cookie (required) and optionally set up Feishu webhook and other environment variables. Start the service with lanhu-mcp --transport http or use run-stdio.sh for stdio mode.
License and pricing
The project is open source under the MIT license. Pricing is not stated; the repository is marked as open source.
Based on the project's public README and repository metadata.