关于
This skill provides a comprehensive, four-phase workflow for architecting, implementing, and evaluating MCP servers using TypeScript or Python. It standardizes the development process by offering specific patterns for tool naming, input/output schemas using Zod or Pydantic, and context management to ensure agents can discover and use tools effectively. By following this guide, developers can build robust integrations that feature actionable error handling, efficient pagination, and rigorous evaluation suites to verify LLM performance across complex real-world tasks.
