Yuque MCP Server: Bridge Your Knowledge Base to AI Workflows

Stop context-switching between your Yuque knowledge base and AI conversations. This MCP server transforms your Yuque documents into structured, AI-accessible content that flows directly into your development workflow.

The Knowledge Access Problem

Your team's best practices, API documentation, and architectural decisions live in Yuque. But when you're coding with AI assistance, that knowledge stays trapped in browser tabs. You end up copying-pasting documentation, explaining context repeatedly, or worse - working with incomplete information.

Direct Knowledge Integration

This MCP server creates a live bridge between your Yuque workspace and MCP-compatible AI clients. Your AI assistant gets instant access to your team's knowledge base without you manually feeding it context every single conversation.

Key Benefits:

• Zero Context Loss: AI assistants see your latest documentation, not stale screenshots or copied text
• Real-time Updates: Webhook integration means changes in Yuque appear immediately in AI conversations
• Structured Access: Documents come through as properly formatted JSON, not raw HTML dumps
• Memory-Efficient Caching: Built-in 5-minute cache prevents API hammering while keeping content fresh

Perfect for These Workflows

API Development: AI assistants can reference your latest API specs stored in Yuque when helping you write integration code, generate tests, or debug endpoint issues.

Code Reviews: Pull your coding standards and architectural guidelines directly into review discussions. No more "check the wiki" comments.

Onboarding Automation: New team members get AI assistance that actually knows your codebase conventions, deployment procedures, and project history.

Documentation-Driven Development: Write specs in Yuque, then have AI generate boilerplate code that actually follows your documented patterns.

Quick Integration

The server runs as a standard HTTP service alongside your existing tools:

# Clone and install
git clone https://github.com/HenryHaoson/Yuque-MCP-Server.git
cd Yuque-MCP-Server && npm install

# Configure with your Yuque token (needs docs:read scope)
echo "YUQUE_TOKEN=your_token_here" > .env

# Build and run
npm run build && npm start

The server exposes your Yuque content through standard MCP endpoints that any compatible AI client can consume. Documents flow through as structured JSON with proper metadata, making them much more useful than raw text dumps.

Smart Caching and Updates

Built-in memory caching (configurable TTL) means your AI assistant gets fast responses without overwhelming Yuque's API. The webhook endpoint lets Yuque push real-time updates, so your AI always works with current information.

No complex authentication flows or permission management - just a simple Yuque Personal Access Token and you're connected.

Why This Matters

Knowledge bases are only valuable when they're accessible at the moment you need them. This server eliminates the friction between where your knowledge lives and where your development work happens. Your AI assistant becomes genuinely helpful because it knows what you know.

Get it running in under 5 minutes and start building with AI that actually understands your project context.