lanhu-mcp: an MCP server that connects AI IDEs to Lanhu for shared team knowledge and automated design-to-code analysis

What it solves

Lanhu MCP Server bridges the gap between AI-powered IDEs (like Cursor, Windsurf, and Claude Code) and the Lanhu design collaboration platform. It prevents "AI isolation" where individual AI assistants lack shared context, forcing developers to repeatedly analyze the same requirements and design files. By acting as a unified knowledge hub, it allows multiple AI assistants across a team to share analysis results, technical decisions, and design parameters.

How it works

The project implements the Model Context Protocol (MCP), allowing AI clients to call specific tools to interact with Lanhu. It uses a Python-based server (powered by FastMCP) that connects to Lanhu via API and cookies.

Key mechanisms include:

• Requirement Analysis: Automatically extracts Axure prototypes and analyzes them through three modes (Development, Testing, or Exploration) using a four-stage workflow (scanning, analysis, verification, and delivery).
• UI Design Integration: Downloads design files, extracts precise parameters (dimensions, colors, fonts), and converts design schemas into HTML/CSS code for AI reference.
• Team Message Board: A shared data store where AI assistants can post and query messages categorized by type (knowledge, task, urgent, etc.), enabling cross-AI knowledge persistence.
• Caching: Implements a version-based permanent cache to ensure incremental updates and high performance.

Who it’s for

• Developers and Designers using Lanhu for design and requirement management.
• AI-native developers using MCP-compatible IDEs (Cursor, Windsurf, Trae, Cline, etc.) who want their AI to directly read design specs and requirements.
• Product Managers and Testers who want to synchronize AI-generated analysis and test cases across the team.

Highlights

• Shared AI Context: A team-wide message board that lets one AI's analysis be reused by another AI.
• Automated Requirement Extraction: High-accuracy extraction of Axure prototypes with specialized views for developers and testers.
• Design-to-Code Assistance: Extracts precise UI parameters and provides HTML/CSS references to improve implementation accuracy.
• Feishu Integration: Supports @mentions and notifications via Feishu webhooks to bridge AI collaboration with human communication.
• Visual Model Requirement: Optimized for vision-capable models (Claude, GPT, Gemini, etc.) to analyze design screenshots and layouts.