01Support for Python (FastMCP) and Node/TypeScript SDK implementations
02Structured 4-phase workflow from deep research to evaluation
03Automated evaluation framework for testing tool performance
04Agent-centric tool design principles for optimized AI workflows
05Best practices for context window management and error handling
0610 GitHub stars