Summary Lanhu MCP Server tackles AI-assisted development fragmentation via a central Model Context Protocol (MCP) server. It enables AI assistants to share team knowledge, analyze requirements from tools like Axure, and extract UI assets, boosting efficiency. Target audience: development teams, product managers, testers, and designers using AI tools (Cursor, Windsurf, Claude Code).

How It Works The server integrates with design platforms (Lanhu, Axure) via MCP, acting as a unified knowledge hub. It uses visual AI models for automated requirement analysis, UI asset extraction with semantic naming, and a shared "message board" for team context. Key technologies: FastMCP, version-based caching, incremental updates, concurrency. This breaks AI IDE silos, allowing AI assistants to access and reuse information, preventing redundant work and preserving knowledge.

Quick Start & Requirements Installation: AI-assisted setup (Cursor) or Docker (docker-compose up -d). Manual source install (easy-install.sh or pip install -r requirements.txt) also supported. Prerequisites: Python 3.10+, visual AI models (Claude, GPT, Gemini, Qwen); text-only models not supported. Lanhu account/Cookie mandatory. Playwright Chromium needed for source installs. Essential config: Lanhu Cookie. Optional: Feishu webhooks, role mappings. Links to docs/contributions available.

Highlighted Details

• 95% accuracy in requirement analysis (dev, test, explore modes).

• "Team Message Board" for shared context, breaking AI IDE isolation.
• Automates UI asset extraction and semantic naming.
• Intelligent caching and incremental updates for performance.
• Integrates with popular AI tools (Cursor, Windsurf, Claude Code) via MCP.

Maintenance & Community Contribution guidelines and issue tracker provided. Roadmap outlines future development.

Licensing & Compatibility MIT License: broad use/modification. Designed for AI tools and Lanhu integration.

Limitations & Caveats Requires visual AI models; text-only incompatible. Valid Lanhu Cookie essential. Third-party, community-developed, not official Lanhu. Users manage credentials securely and assume usage risks.