Fathom MCP Server
Enables AI agents to access Fathom AI meeting data including transcripts, summaries, teams, and team members. Provides tools to list meetings with filters, retrieve detailed transcripts and summaries, and manage team information.
README
Fathom MCP Server
A Model Context Protocol (MCP) server that provides AI agents with access to Fathom AI meeting data, including transcripts, summaries, teams, and team members.
Features
This MCP server exposes 5 tools for interacting with Fathom AI:
fathom_list_meetings- List meetings with optional filters (date ranges, invitees, teams, etc.)fathom_get_summary- Get meeting summary by recording IDfathom_get_transcript- Get meeting transcript with speaker information and timestampsfathom_list_teams- List all teams in the organizationfathom_list_team_members- List team members for a specific team
Architecture
- Framework: Next.js 15 with App Router
- MCP Adapter: mcp-handler (Vercel's official MCP adapter)
- Transport: Supports both Streamable HTTP and Server-Sent Events (SSE)
- Deployment: Vercel with automatic HTTPS and global CDN
- Type Safety: Full TypeScript support with Zod validation
Quick Start
1. Clone and Install
git clone <your-repo-url>
cd fathom-mcp-server
npm install
2. Environment Setup
Copy the environment template and add your Fathom API key:
cp .env.example .env.local
Edit .env.local:
FATHOM_API_KEY=your_fathom_api_key_here
Get your API key from Fathom Settings.
3. Local Development
npm run dev
The MCP server will be available at http://localhost:3000/api/mcp
4. Deploy to Vercel
Or deploy manually:
npm install -g vercel
vercel
Add your FATHOM_API_KEY environment variable in the Vercel dashboard.
Client Integration
OpenAI AgentKit
Connect directly to the HTTP endpoint with Bearer token authentication:
{
"mcpServers": {
"fathom": {
"url": "https://fathom-mcp.vercel.app/api/mcp",
"auth": {
"type": "bearer",
"token": "YOUR_FATHOM_API_KEY"
}
}
}
}
Configuration Steps:
- Replace
YOUR_FATHOM_API_KEYwith your actual Fathom AI API key - The API key is passed as a Bearer token in the Authorization header
- This keeps the API key client-side controlled and secure
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"fathom": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://fathom-mcp.vercel.app/api/mcp"
]
}
}
}
Config file locations:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\\Claude\\claude_desktop_config.json
Cursor
Add to your ~/.cursor/mcp.json:
{
"mcpServers": {
"fathom": {
"url": "https://your-app.vercel.app/api/mcp"
}
}
}
Tool Usage Examples
List Recent Meetings
{
"name": "fathom_list_meetings",
"arguments": {
"created_after": "2024-01-01T00:00:00Z",
"include_summary": true,
"limit": 10
}
}
Get Meeting Transcript
{
"name": "fathom_get_transcript",
"arguments": {
"recording_id": 123456789
}
}
Filter Meetings by Team
{
"name": "fathom_list_meetings",
"arguments": {
"teams": ["Engineering", "Product"],
"include_action_items": true
}
}
List Team Members
{
"name": "fathom_list_team_members",
"arguments": {
"team_id": "team_123"
}
}
API Reference
fathom_list_meetings
List meetings with optional filters.
Parameters:
calendar_invitees(string[]): Email addresses to filter bycalendar_invitees_domains(string[]): Company domains to filter bycalendar_invitees_domains_type(enum): Filter by external domainscreated_after(string): ISO 8601 timestampcreated_before(string): ISO 8601 timestampinclude_transcript(boolean): Include transcript datainclude_summary(boolean): Include summary datainclude_action_items(boolean): Include action itemsinclude_crm_matches(boolean): Include CRM matcheslimit(number): Max results (1-100)cursor(string): Pagination cursorrecorded_by(string[]): Filter by recorder emailsteams(string[]): Filter by team names
fathom_get_summary
Get meeting summary by recording ID.
Parameters:
recording_id(number): The recording ID
fathom_get_transcript
Get meeting transcript with speaker information.
Parameters:
recording_id(number): The recording ID
fathom_list_teams
List all teams in the organization.
Parameters:
cursor(string): Optional pagination cursor
fathom_list_team_members
List team members for a specific team.
Parameters:
team_id(string): The team IDcursor(string): Optional pagination cursor
Error Handling
The server includes comprehensive error handling:
- Missing API Key: Clear error message if
FATHOM_API_KEYis not set - Fathom API Errors: Properly formatted error responses with status codes
- Rate Limiting: Graceful handling of 429 responses
- Input Validation: Zod schemas validate all tool inputs
- Network Errors: Timeout and connection error handling
Security Considerations
- HTTPS Only: Vercel provides automatic TLS certificates
- CORS Headers: Configured for OpenAI's infrastructure
- API Key Security: Stored in environment variables, never committed
- Input Validation: All inputs validated with Zod schemas
- Rate Limiting: Respects Fathom API rate limits
Development
Project Structure
├── app/
│ └── api/
│ └── [transport]/
│ └── route.ts # MCP route handler
├── lib/
│ └── fathom-client.ts # Fathom API client
├── .env.example # Environment template
├── vercel.json # Vercel configuration
└── README.md # This file
Testing
Test the MCP server locally:
# Start the development server
npm run dev
# Test with curl
curl -X POST http://localhost:3000/api/mcp \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/list"
}'
Type Checking
npm run type-check
Troubleshooting
Common Issues
-
"FATHOM_API_KEY environment variable is required"
- Ensure you've set the environment variable in
.env.local(local) or Vercel dashboard (production)
- Ensure you've set the environment variable in
-
"Fathom API Error (401)"
- Check that your API key is valid and has the correct permissions
-
"Fathom API Error (429)"
- You've hit Fathom's rate limit. Wait before making more requests
-
CORS errors in browser
- The server includes proper CORS headers. If issues persist, check your client configuration
Debug Mode
Enable verbose logging in development:
NODE_ENV=development npm run dev
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests if applicable
- Submit a pull request
License
MIT License - see LICENSE file for details.
Support
- Fathom API Documentation: developers.fathom.ai
- MCP Specification: modelcontextprotocol.io
- Issues: Create an issue in this repository
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。