Fathom AI MCP Server

Fathom AI MCP Server

Enables AI assistants to access Fathom meeting recordings, summaries, transcripts, teams, and webhooks through the Model Context Protocol.

Category
访问服务器

README

Fathom AI MCP Server

License: MIT Python 3.10+ FastMCP

A Model Context Protocol (MCP) server for interacting with the Fathom AI API. This server provides tools for accessing meeting recordings, summaries, transcripts, teams, and webhooks.

Quick Start

git clone https://github.com/Dot-Fun/fathom-mcp.git
cd fathom-mcp
uv pip install fastmcp httpx pydantic
export FATHOM_API_KEY="your_api_key_here"
fastmcp run server.py

Features

Tools

  • list_meetings: List meetings with advanced filtering (by participants, date ranges, teams) and optional inclusion of transcripts, summaries, action items, and CRM matches
  • get_summary: Retrieve meeting summaries (supports async delivery)
  • get_transcript: Retrieve meeting transcripts with speaker information and timestamps (supports async delivery)
  • list_teams: List all accessible teams
  • list_team_members: List team members with optional team filtering
  • create_webhook: Create webhooks for meeting notifications with customizable triggers and content inclusion
  • delete_webhook: Delete webhooks by ID

Resources

  • fathom://api/info: API information and available endpoints
  • fathom://api/rate-limits: Rate limiting information and best practices

Installation

  1. Clone this repository or copy the files to your local machine

  2. Install dependencies using uv (recommended) or pip:

# Using uv (recommended)
uv pip install -e .

# Or using pip
pip install -e .
  1. Set up your Fathom API key:

Create a .env file in the project root:

FATHOM_API_KEY=your_api_key_here

The server auto-loads .env for local usage.

Get your API key from the Fathom settings page.

Usage

Running the Server

Local Development (stdio)

# Using fastmcp CLI
fastmcp run server.py

# Or directly with Python
python server.py

HTTP Server

# Modify server.py to run as HTTP server
if __name__ == "__main__":
    mcp.run(transport="http", host="0.0.0.0", port=8000)

Integrating with Claude Desktop

Add to your Claude Desktop configuration (claude_desktop_config.json):

{
  "mcpServers": {
    "fathom": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/fathom-mcp",
        "run",
        "fastmcp",
        "run",
        "server.py"
      ],
      "env": {
        "FATHOM_API_KEY": "your_api_key_here"
      }
    }
  }
}

Integrating with Codex (alternative)

Add this block to your Codex config (~/.codex/config.toml):

[mcp_servers.fathom]
command = "uv"
args = ["--directory", "/path/to/fathom-mcp", "run", "fastmcp", "run", "server.py"]

Then restart Codex (or reload MCP servers) and verify:

codex mcp list

API Reference

Authentication

All requests require the FATHOM_API_KEY environment variable. API keys are user-level and can access:

  • Meetings recorded by the user
  • Meetings shared to the user's team

Rate Limits

  • Global limit: 60 API calls per 60-second window
  • Rate-limited requests return HTTP 429
  • Monitor rate limit headers: RateLimit-Limit, RateLimit-Remaining, RateLimit-Reset

Tool Examples

List Recent Meetings

# List meetings with transcripts from the last week
result = await list_meetings(
    created_after="2024-01-01T00:00:00Z",
    include_transcript=True,
    include_summary=True
)

Get Meeting Summary

# Get summary for a specific recording
summary = await get_summary(recording_id=123456789)

Create a Webhook

# Create webhook for team recordings with all content
webhook = await create_webhook(
    destination_url="https://your-app.com/webhook",
    triggered_for=["my_recordings", "shared_team_recordings"],
    include_transcript=True,
    include_summary=True,
    include_action_items=True,
    include_crm_matches=True
)
# Returns webhook ID, URL, and secret for signature verification

Filter Meetings by Participants

# Find meetings with specific participants
meetings = await list_meetings(
    calendar_invitees=["alice@acme.com", "bob@acme.com"],
    include_action_items=True
)

List Team Members

# Get all members of the Sales team
members = await list_team_members(team="Sales")

Error Handling

The server handles common API errors:

  • 401 Unauthorized: Invalid or missing API key
  • 400 Bad Request: Invalid parameters
  • 404 Not Found: Resource doesn't exist
  • 429 Rate Limited: Too many requests (includes reset time)

Development

Project Structure

fathom-mcp/
├── server.py           # Main MCP server implementation
├── pyproject.toml      # Project configuration and dependencies
├── Dockerfile          # Container configuration for deployment
├── package.json        # Package metadata
├── README.md           # This file
├── .env                # Environment variables (API key)
└── .dockerignore       # Docker build exclusions

Code Quality

The project uses Ruff for linting and formatting:

# Install dev dependencies
uv pip install -e ".[dev]"

# Run linter
ruff check .

# Auto-fix issues
ruff check --fix .

# Format code
ruff format .

Testing

The server has been tested with real Fathom API data. See TEST_RESULTS.md for detailed test results.

To test manually:

# Set your API key
export FATHOM_API_KEY="your_api_key_here"

# Test server loads
python -c "from server import mcp; print('✓ Server ready')"

# Test API connectivity
python -c "
import asyncio
from server import make_request
asyncio.run(make_request('GET', '/meetings'))
"

API Documentation

For complete API documentation, visit:

License

This project is open source and available under the MIT License.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

Acknowledgments

Built with FastMCP - The fast, Pythonic way to build MCP servers.

Repository

GitHub: https://github.com/Dot-Fun/fathom-mcp

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选