|
Build MCP servers in Python with structured tools, resources, and prompts exposed to LLMs. Supports three server modes: local stdio, HTTP transport, and FastMCP Cloud deployment with module-level server export Includes 8 built-in middleware types (timing, caching, rate limiting, error handling, logging) with configurable execution order and custom middleware support Provides 4 authentication patterns: token validation (JWTVerifier), external OAuth providers (RemoteAuthProvider), OAuth Proxy for GitHub/Google/Azure/AWS/Discord, and full authorization servers Offers background tasks with progress tracking (task=True), agentic sampling with tools (ctx.sample), and context-based elicitation for user input Supports 6+ storage backends (memory, disk, Redis, DynamoDB, MongoDB) with Fernet encryption for OAuth tokens and persistent state; prevents 30+ common errors including lifespan misconfiguration, middleware ordering, and resource URI mismatches FastMCP - Build MCP Servers in Python FastMCP is a Python framework for building Model Context Protocol (MCP) servers that expose tools, resources, and prompts to Large Language Models like Claude. This skill provides production-tested patterns, error prevention, and deployment strategies for building robust MCP servers. Quick Start Installation pip install fastmcp # or uv pip install fastmcp Minimal Server from fastmcp import FastMCP
don't have the plugin yet? install it then click "run inline in claude" again.