FastMCP Development

This skill provides an end-to-end development environment for building Model Context Protocol (MCP) servers using the FastMCP framework. It is designed for software engineers and developers looking to create LLM-executable functions that bridge internal systems with AI assistants like Claude. By leveraging FastMCP, users can rapidly transform standard Python or TypeScript functions into highly functional MCP tools, resources, and workflows without manually handling complex protocol serialization.

• Streamlines the creation of MCP servers with structured, step-by-step development guidance.

• Supports both Python and TypeScript, offering clear templates and best practice examples for each.

• Facilitates the definition of LLM tools (functions), dynamic data resources, and instructional prompts for complex task orchestration.

• Provides robust instructions for integrating servers with Claude Desktop via STDIO or HTTP/SSE transports.

• Includes security-focused guidance for production environments, including OAuth implementation and token verification.

• Offers diagnostic steps for testing server connectivity and verifying end-to-end integration between the MCP server and the AI agent.

• Follow the mandatory 10-step Todo list provided in the skill to ensure all production requirements, such as documentation and validation, are met.

• Utilize the provided code snippets as a reference for type hints, Pydantic field validation, and context injection.

• Use STDIO mode for rapid local development and iteration, switching to HTTP/SSE for persistent, network-accessible infrastructure.

• Consult the documentation for specific library versions and dependency requirements to maintain compatibility with the latest FastMCP v3 specifications.

• Ensure all tools intended for AI execution include descriptive docstrings and proper parameter schemas for optimal LLM function calling performance.