Skip to content

README.md

Latest commit

 

History

History
103 lines (77 loc) · 3.81 KB

README.md

File metadata and controls

103 lines (77 loc) · 3.81 KB

MCP Doctor Documentation

Welcome to the comprehensive documentation for MCP Doctor - the diagnostic tool for MCP (Model Context Protocol) servers.

📚 Documentation Index

Core Concepts

Features

  • Tool Description Analysis - Static analysis of tool descriptions and parameters
  • Token Efficiency Analysis - Dynamic testing of tool response sizes and optimization
  • CLI Interface - Command-line usage and options
  • Python API - Programmatic usage in Python applications

Advanced Topics

  • Architecture Overview - How MCP Doctor works internally
  • Custom Checkers - Creating your own diagnostic modules
  • Performance Optimization - Tips for faster analysis
  • Security Considerations - Safe handling of sensitive data
  • Code Coverage - Test coverage measurement and reporting

Troubleshooting

  • Token Efficiency Troubleshooting - Solutions for token efficiency analysis issues
  • Common Issues - Solutions to frequent problems
  • Debugging Guide - How to debug analysis issues
  • FAQ - Frequently asked questions

🚀 Quick Start

# Install MCP Doctor
pip install mcp-doctor

# Analyze tool descriptions
mcp-doctor analyze --target "npx your-mcp-server"

# Analyze token efficiency
mcp-doctor analyze --target "npx your-mcp-server" --check token_efficiency

# Run all available checks
mcp-doctor analyze --target "npx your-mcp-server" --check all

🎯 Key Features

📝 Tool Description Analysis

Evaluates MCP tools for AI agent compatibility:

  • Description clarity and completeness
  • Parameter naming conventions
  • Usage context and examples
  • Technical jargon detection

🔢 Token Efficiency Analysis

Measures actual tool performance:

  • Response size measurement (following Anthropic's 25k token guideline)
  • Pagination and filtering capability detection
  • Verbose identifier flagging
  • Performance metrics collection

🏗️ Architecture

MCP Doctor is built with a modular architecture:

mcp-doctor/
├── src/mcp_analyzer/
│   ├── cli.py              # Command-line interface
│   ├── mcp_client.py       # MCP server communication
│   ├── mcp_stdio_client.py # STDIO transport
│   ├── mcp_sse_client.py   # SSE transport
│   ├── reports.py          # Output formatting
│   └── checkers/           # Diagnostic modules
│       ├── descriptions.py      # Description analysis
│       └── token_efficiency.py  # Token efficiency analysis
└── tests/                  # Test suite

🤝 Contributing

MCP Doctor is open source and welcomes contributions:

  1. Bug Reports - Found an issue? Please report it
  2. Feature Requests - Have an idea? We'd love to hear it
  3. Code Contributions - Pull requests are welcome
  4. Documentation - Help improve these docs

🔗 Related Resources

📞 Support

Built with ❤️ by Destilabs for the AI agent development community