A hands-on learning project for Model Context Protocol (MCP) development, progressing from basic servers to advanced enterprise features.
Model Context Protocol enables LLMs to securely access external data sources and tools, creating a standardized way for AI models to interact with various resources and services.
learnmcp/
├── LEARNING_PLAN.md # Detailed learning roadmap
├── phase1-basic-server/ # Foundation: Basic MCP server setup
├── phase2-resources-prompts/ # Core: Resources and prompts management
├── phase3-integration/ # Practical: Real-world API integration
└── phase4-advanced/ # Enterprise: Advanced features and patterns
- Choose your starting phase: Begin with
phase1-basic-server/ - Follow the learning plan: Refer to
LEARNING_PLAN.mdfor detailed guidance - Test your servers: Use MCP inspector to test your implementations
- Document your learnings: Add notes and examples as you progress
- Node.js 18+
- TypeScript
- Git
# Clone or navigate to this project
cd phase1-basic-server
# Install dependencies (when we add them)
npm install
# Run the server
npm start- MCP server architecture and lifecycle
- Tool creation and management
- Resource and prompt handling
- API integration patterns
- Authentication and security
- Deployment and monitoring
This is a personal learning project. Feel free to experiment and modify examples to deepen your understanding of MCP concepts.