MCP 服务器

FastAPI Docs MCP Server

Provides AI assistants with real-time access to FastAPI documentation, enabling search, browsing, code examples, and comparisons.

README

<h1 align="center">🐍 FastAPI Docs MCP Server 🤖</h1>

<p align="center"> <em>Real-time FastAPI documentation access for AI assistants</em> </p> <p align="center"> <a href="https://github.com/jaredthivener/fastapi-docs-mcp/actions/workflows/test.yml" target="_blank"><img src="https://github.com/jaredthivener/fastapi-docs-mcp/actions/workflows/test.yml/badge.svg?event=push&branch=main" alt="Test"></a> <a href="https://github.com/jaredthivener/fastapi-docs-mcp/actions/workflows/codeql.yml" target="_blank"><img src="https://github.com/jaredthivener/fastapi-docs-mcp/actions/workflows/codeql.yml/badge.svg" alt="CodeQL"></a> <a href="https://github.com/jaredthivener/fastapi-docs-mcp/blob/main/LICENSE" target="_blank"><img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="License"></a> <a href="https://github.com/jaredthivener/fastapi-docs-mcp" target="_blank"><img src="https://img.shields.io/badge/python-3.13%2B-blue.svg" alt="Python 3.13+"></a> </p>

An MCP (Model Context Protocol) server that provides real-time access to FastAPI documentation. Use it with Claude, GitHub Copilot, or any MCP-compatible client to instantly query FastAPI docs.

Features

Real-time documentation — Fetches directly from fastapi.tiangolo.com
Smart search — Find docs by keyword with common alias support
Full sitemap access — Browse all available documentation pages
Code examples — Get just the code, no prose
Compare approaches — Side-by-side comparisons of different patterns

Tools

Tool	Description
`get_fastapi_docs(path)`	Fetch any documentation page by path
`search_fastapi_docs(query)`	Search docs by keyword (with alias support)
`list_fastapi_pages()`	List all available documentation pages
`get_fastapi_example(topic)`	Get just the code examples, no prose
`compare_fastapi_approaches(topic)`	Compare different approaches side-by-side
`get_fastapi_best_practices(topic)`	Get combined best practices from multiple pages

Examples

Once connected, ask your AI assistant:

"How do I set up CORS in FastAPI?"
"Show me the FastAPI security documentation"
"What are FastAPI dependencies?"
"List all FastAPI tutorial pages"
"Give me a code example for JWT authentication"
"Compare sync vs async in FastAPI"
"What are the best practices for testing in FastAPI?"

Installation

git clone https://github.com/jaredthivener/fastapi-docs-mcp.git
cd fastapi-docs-mcp
uv sync

Usage

With Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "fastapi-docs": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/fastapi-docs-mcp", "python", "main.py"]
    }
  }
}

With VS Code (GitHub Copilot)

Add to .vscode/mcp.json in your workspace:

{
  "servers": {
    "fastapi-docs": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/fastapi-docs-mcp", "python", "main.py"]
    }
  }
}

With Docker

Build the image locally:

docker build -t fastapi-docs-mcp .

Then use this MCP server config:

{
  "servers": {
    "fastapi-docs": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "fastapi-docs-mcp"
      ],
      "type": "stdio"
    }
  }
}

Standalone

uv run python main.py

Development

# Install dev dependencies
uv sync --extra dev

# Run tests
uv run pytest

# Lint
uv run ruff check .

# Format
uv run ruff format .

# Type check
uv run mypy main.py

How It Works

The server fetches documentation directly from the official FastAPI website:

Sitemap-based discovery — Uses sitemap.xml to find all available pages
Real-time fetching — Retrieves current documentation on each request
Smart extraction — Extracts readable content from HTML pages
Keyword aliases — Maps common terms (e.g., "auth" → "security")

Content Limits

To keep responses fast and avoid overloading AI context windows, documentation content is truncated to a maximum length (currently 15,000 characters). If a page exceeds that limit, the response ends with a truncation notice and a link to the full page.