MCP Development Environment

A comprehensive Docker-based development environment for building and testing Model Context Protocol (MCP) servers.

🎯 What's Included

Core Services:

• Python MCP Server (Port 8000) - Full-featured implementation with debugging

• Node.js MCP Server (Port 3000) - Alternative implementation

• PostgreSQL (Port 5432) - Database with sample data

• Redis (Port 6379) - Caching layer

• Nginx File Server (Port 8080) - Static file serving with CORS

Development Features:

• Hot reloading for both Python and Node.js

• Built-in debugger support (Python: 5678, Node.js: 9229)

• Comprehensive logging and monitoring

• Pre-configured testing frameworks

• Sample data and schemas

• Code quality tools (linting, formatting, type checking)

Related MCP server: Docker MCP Server

🚀 Quick Start

1. Setup the environment:

# Make the setup script executable and run it
chmod +x setup.sh
./setup.sh

2. Verify everything is working:

# Check service status
docker-compose ps

# Test endpoints
curl http://localhost:8080/health  # File server
curl http://localhost:8000/health  # Python MCP server  
curl http://localhost:3000/health  # Node.js MCP server

3. Start developing:

# Edit the Python MCP server
vim src/main.py

# View logs in real-time
docker-compose logs -f mcp-server

# Run tests
docker-compose exec mcp-server python -m pytest

🔧 Key Features of the MCP Servers

Available Tools:

• write_file - Write content to files

• execute_sql - Run database queries

• cache_set/get - Redis cache operations

• list_directory - Browse file system

• analyze_data - Basic data analysis on CSV files

Resources:

• File system access to /data directory

• Database table schemas and sample data

• Configuration files and documentation

Sample Usage:

# The Python server provides tools for:
await mcp_server.call_tool("write_file", {
    "path": "analysis.txt", 
    "content": "Sample analysis results"
})

await mcp_server.call_tool("execute_sql", {
    "query": "SELECT * FROM users WHERE department = $1",
    "parameters": ["Engineering"] 
})

🐛 Debugging Setup

Python (VSCode):

{
  "name": "Python: Remote Attach",
  "type": "python", 
  "request": "attach",
  "connect": {"host": "localhost", "port": 5678},
  "pathMappings": [
    {"localRoot": "${workspaceFolder}/src", "remoteRoot": "/app/src"}
  ]
}

Node.js (Chrome DevTools):

• Open chrome://inspect

• Connect to localhost:9229

📊 Monitoring & Logs

# View all service logs
docker-compose logs -f

# Monitor specific service
docker-compose logs -f mcp-server

# Check resource usage
docker stats

# Database operations
docker-compose exec postgres psql -U mcp_user -d mcp_dev

# Redis operations  
docker-compose exec redis redis-cli

🛠️ Development Workflow

The environment supports both transport methods:

• stdio (default) - For direct MCP client integration

• HTTP/WebSocket - For web-based development and testing

You can easily switch between implementations or run both simultaneously for comparison and testing.

🗂️ Project Structure

mcp/
├── src/                    # Python MCP server source
│   └── main.py            # Main Python server implementation
├── src-node/              # Node.js MCP server source
│   └── server.js          # Main Node.js server implementation
├── db/                    # Database initialization scripts
│   ├── init.sql           # Schema and tables
│   └── sample_data.sql    # Sample data
├── data/                  # Data files (mounted to containers)
├── static/                # Static files served by Nginx
├── tests/                 # Test suites
├── .vscode/               # VSCode debug configuration
├── docker-compose.yml     # Service definitions
├── python.Dockerfile      # Python server container
├── node.Dockerfile        # Node.js server container
├── nginx.conf             # Nginx configuration
├── setup.sh               # Setup and management script
└── README.md              # This file

🔧 Management Commands

The setup.sh script provides convenient management:

./setup.sh setup     # Initial setup and start (default)
./setup.sh start     # Start services
./setup.sh stop      # Stop services  
./setup.sh restart   # Restart services
./setup.sh status    # Show service status
./setup.sh logs      # Show service logs
./setup.sh clean     # Remove everything (with confirmation)
./setup.sh help      # Show help

🧪 Testing

Both Python and Node.js servers include comprehensive test suites:

# Run Python tests
docker-compose exec mcp-server python -m pytest tests/ -v

# Run Node.js tests  
docker-compose exec mcp-server-node npm test

# Run tests with coverage
docker-compose exec mcp-server python -m pytest tests/ --cov=src

🔍 Database Schema

The PostgreSQL database includes several sample tables:

• users - User accounts with departments and roles

• products - Product catalog with categories and inventory

• orders - Order history with status tracking

• order_items - Order line items

• analytics_events - Event tracking data

• app_config - Application configuration

📡 API Endpoints

File Server (Port 8080):

• GET /health - Health check

• GET /data/ - Browse data directory

• GET /static/ - Browse static files

• GET /api/docs - API documentation

Python MCP Server (Port 8000):

• GET /health - Health check

• MCP protocol via stdio transport

Node.js MCP Server (Port 3000):

• GET /health - Health check

• MCP protocol via stdio transport

🚨 Troubleshooting

Services not starting:

1. Check Docker is running: docker info

2. Check port conflicts: netstat -tulpn | grep :8000

3. View startup logs: docker-compose logs

Database connection issues:

# Test database connectivity
docker-compose exec postgres pg_isready -U mcp_user

# Connect to database manually
docker-compose exec postgres psql -U mcp_user -d mcp_dev

Redis connection issues:

# Test Redis connectivity
docker-compose exec redis redis-cli ping

Debug not working:

• Ensure debug ports (5678, 9229) are not in use

• Check firewall settings

• Verify VSCode debug configuration matches container setup

🤝 Contributing

1. Fork the repository

2. Make changes in your environment

3. Test thoroughly with provided test suites

4. Submit a pull request

📄 License

This project is provided as-is for development and testing purposes.

This environment gives you a complete MCP development platform with real databases, caching, file systems, and debugging tools - perfect for building and testing production-ready MCP servers!