dsphper/lanhu-mcp

This page was translated automatically. Technical details may differ from the original. View the original

2.5K stars 277 forks Quality 90 Trend 52MIT

⚡ Requirements analysis efficiency increased by 200%! The world's first team-collaborative MCP server designed for the AI programming era, automatically analyzing requirements and writing front- and backend code, downloading sliced images

Overview

A powerful Model Context Protocol (MCP) server, designed specifically for the AI programming era, perfectly supports the Lanhu design collaboration platform. 🔥 : - 📋 : Automatically extracts Axure prototypes, three analysis modes (development/testing/exploration), requirements analysis accuracy >95% - 💬 : Breaks down AI IDE silos, allowing all AI assistants to share knowledge bases and contexts - 🎨 : Automatically downloads design drafts, intelligently extracts slices, semantic naming; Design drawing analysis can obtain precise parameters such as size/spacing/color/font, and provides converted HTML+CSS code references - ⚡ : Intelligent caching based on version numbers, incremental updates, and concurrent processing 🎯 : - ✅ Cursor + Bluelake: Let Cursor AI directly read Bluelake requirements documents and design drafts - ✅ Windsurf + Bluelake: Windsurf Cascade AI directly reads Bluelake requirements documents and design drafts - ✅ Claude Code + Bluelake: Claude AI reads Bluelake requirements documents and design drafts directly - ✅ OpenClaw + Bluelake: OpenClaw natively supports reading Bluelake requirements documents and design drafts - ✅ ClawBot + Bluelake: ClawBot intelligent assistant deeply integrates Bluelake collaboration - ✅ Trae + Bluelake: Trae AI directly reads Bluelake requirements documents and design drafts - ✅ Tongyi Lingma + Bluelake: Tongyi Lingma AI directly reads Bluelake requirements documents and design drafts - ✅ Cline + Bluelake: Cline AI reads Bluelake requirements documents and design drafts directly - ✅ any AI development tool that supports the MCP protocol 🎯 : - ❌ : Each developer's AI works independently, repeatedly analyzes requirements, and cannot share experience - ✅ : All AIs are connected to the same knowledge hub, needs analysis is conducted once, everyone reuses them, and experiences encountered are permanently preserved - : Automatically download and parse all pages, resources, and interactions of Axure prototypes - : 🔧 : Detailed field rules, business logic, global flowcharts - 🧪 : Test scenarios, use cases, boundary values, validation rules - 🚀 : Core feature overview, module dependencies, review points - : Global scanning → Group analysis → Reverse validation → Generate deliverables - : TODO-driven systematic analysis flow - :Batch download and display UI design drawings - :During analysis, not only does it return to the design preview, but it also retrieves **(component size, spacing, color values, font size, etc.), and automatically converts the design schema to match the native Bluelake export effect, facilitating AI reference implementation. - :Automatically recognizes and exports design slices and icon resources-...

Overview (machine translated)

Enabling All AI Assistants to Share Team Knowledge, Breaking AI IDE Silos

English | Chinese Simplified

Quick Start • Features • Documentation • Contribution Guide

🌟 Project Highlights

A Model Context Protocol (MCP) server for Bluelake design delivery and requirement reading. MCP provides source data and resources, and the large model combines visuals to understand and adapt to target projects.

**v1.8.7: Local stdio now includes socksio and an optional direct connection switch; Added a Docker-free local boot guide. ** Version Notes · LOCAL-STDIO · Design workflows

**v1.8.6: Requirements pages support precise partial drawing by text, stable block ID, block ID, or coordinate; Long page segments remain complete horizontally. ** Version notes

🔥 Core Innovations:

📋 Requirements Analysis Support: Extract Axure pages, text, and comments to provide perspectives for development, testing, and exploration; Business conclusions are checked by AI and users.
💬 Team Knowledge Base: Break down AI IDE silos, allowing all AI assistants to share knowledge bases and context
🎨 UI Design Support: Automatically downloads design drafts, intelligently extracts sliced images, and semantic naming; Design drawing analysis can obtain precise parameters such as size/spacing/color/font, and provides converted HTML+CSS code references
⚡ Performance Optimization: Intelligent caching based on version numbers, incremental updates, and concurrent processing

🎯 Applicable Scenarios:

✅ Cursor + Blue Lagoon: Allows Cursor AI to directly read Blue Lagoon requirements documents and design drafts
✅ Windsurf + Blue Lake: Windsurf Cascade AI directly reads Blue Lake requirements documents and design drafts
✅ Claude Code + Bluelake: Claude AI directly reads Bluelake requirements documents and design drafts
✅ OpenClaw + Bluelake: OpenClaw natively supports reading Bluelake requirements documents and design drafts
✅ ClawBot + Blue Lake: The ClawBot intelligent assistant deeply integrates Blue Lake collaboration
✅ Trae + Blue Lake: Trae AI directly reads Blue Lake requirements documents and design drafts
✅ Tongyi Lingma + Blue Lake: Tongyi Lingma AI directly reads Blue Lake requirements documents and design drafts
✅ Cline + Bluelake: Cline AI directly reads Bluelake requirements documents and design drafts
✅ Any AI development tool that supports the MCP protocol

🎯 Solving Pain Points:

❌ Old World: Each developer’s AI works independently, repeatedly analyzes requirements, and cannot share experiences
✅ New World: All AIs connect to the same knowledge hub, requiring analysis once, full reuse by all staff, and permanent retention of pitfall experiences

📑 Table of Contents

Core features
Start quickly
Team message board: breaking through the last mile of AI collaboration

This page was translated automatically. Technical details may differ from the original.

Install

This server does not publish a one-line install command.

Open the repository installation guide New to MCP? Start with the mcpskill.org guide →

Configuration

{
  "mcpServers": {
    "lanhu": {
      "type": "http",
      "url": "http://localhost:8000/mcp?role=Developer&name=YourName"
    }
  }
}