Zoom MCP Server
Enables intelligent monitoring and management of Zoom rooms across multiple sites using smart location resolution and hierarchical discovery. It provides tools for checking room status, retrieving location hierarchies, and accessing detailed room configurations through the Zoom API.
README
Zoom MCP Server
A FastMCP Model Context Protocol server that provides intelligent monitoring and management of Zoom rooms across multiple sites with smart location resolution.
🚀 Features (New!)
- 5 Powerful Tools for comprehensive Zoom room management
- Smart Location Resolution with fuzzy matching (e.g., "SF1", "DEN1", "Floor 3")
- Denver Building Aliases - Special hardcoded mappings for room naming compatibility
- Efficient API Usage - Single call for company-wide queries vs. multiple location-specific calls
- OAuth 2.0 Authentication with automatic token refresh and file-based caching
- Hierarchical Location Discovery - Understands campus → building → floor relationships
- User-Friendly Confirmations - Clear messages explaining what was resolved
🛠️ Tools Available
test_zoom_connection
Test Zoom API authentication and connection status.
# Usage: Verify credentials are working
mcp call test_zoom_connection --params '{}' uv run src/server.py
get_zoom_sites
Get all Zoom locations with hierarchy and aliases.
# Usage: Understand available locations and relationships
mcp call get_zoom_sites --params '{}' uv run src/server.py
get_zoom_rooms
Get Zoom rooms with optional smart location filtering.
⚡ IMPORTANT: For maximum efficiency with company-wide queries (e.g., "find offline rooms anywhere"), omit location_query to make a single API call.
# Company-wide (EFFICIENT - single API call)
mcp call get_zoom_rooms --params '{}' uv run src/server.py
# Location-specific (multiple API calls)
mcp call get_zoom_rooms --params '{"location_query":"SF1"}' uv run src/server.py
mcp call get_zoom_rooms --params '{"location_query":"DEN1"}' uv run src/server.py
mcp call get_zoom_rooms --params '{"location_query":"Floor 3"}' uv run src/server.py
get_room_details
Get detailed information about a specific room.
# Usage: Deep dive into specific room configuration
mcp call get_room_details --params '{"room_id":"ROOM_ID_HERE"}' uv run src/server.py
resolve_location
Debug tool to test location resolution without fetching rooms.
# Usage: Debug how location queries get resolved
mcp call resolve_location --params '{"location_query":"DEN2"}' uv run src/server.py
📍 Smart Location Resolution
The server understands various location query patterns:
| Query Pattern | Example | What It Resolves |
|---|---|---|
| Campus codes | SF1, NYC, DEN |
Entire campus with all buildings/floors |
| Building numbers | Building 1, DEN1 |
Specific building or hardcoded alias |
| Floor numbers | Floor 1, 3F |
All floors with that number across sites |
| Partial names | Denver, Francisco |
Best fuzzy match |
Special Denver Building Aliases
Due to Zoom's location hierarchy vs. room naming patterns, Denver has special hardcoded mappings:
DEN1→ Denver Building 1 (Floor 3) → Rooms:DEN-1-101,DEN-1-102, etc.DEN2→ Denver Building 2 (T3F3, T3F5, T3F6) → Rooms:DEN-2-201,DEN-2-202, etc.
🔧 Installation & Setup
Prerequisites
- Python 3.10+
- UV package manager
- Zoom Pro/Business account with API access
1. Clone Repository
git clone https://github.com/chadkunsman/zoom-mcp.git
cd zoom-mcp
2. Install Dependencies
uv pip install -e .
3. Zoom API Configuration
- Create a Server-to-Server OAuth app in Zoom Marketplace
- Add required scope:
room:read:admin - Get your credentials: Account ID, Client ID, Client Secret
4. Configure Credentials
Create .env file:
ZOOM_ACCOUNT_ID=your_account_id_here
ZOOM_CLIENT_ID=your_client_id_here
ZOOM_CLIENT_SECRET=your_client_secret_here
5. Test Installation
# Install MCPTools for testing
brew tap f/mcptools && brew install mcp
# Test the server
mcp tools uv run src/server.py
mcp call test_zoom_connection --params '{}' uv run src/server.py
🔌 MCP Client Configuration
For Claude Desktop and Similar MCP Clients
Add to your MCP client configuration:
Using Environment Variables (Recommended)
{
"mcpServers": {
"zoom-mcp": {
"command": "uv",
"args": [
"run",
"--directory",
"/path/to/zoom-mcp",
"src/server.py"
],
"env": {
"ZOOM_ACCOUNT_ID": "your_account_id_here",
"ZOOM_CLIENT_ID": "your_client_id_here",
"ZOOM_CLIENT_SECRET": "your_client_secret_here"
}
}
}
}
Using Command-Line Arguments
{
"mcpServers": {
"zoom-mcp": {
"command": "uv",
"args": [
"run",
"--directory",
"/path/to/zoom-mcp",
"src/server.py",
"--zoom-account-id",
"your_account_id_here",
"--zoom-client-id",
"your_client_id_here",
"--zoom-client-secret",
"your_client_secret_here"
]
}
}
}
💡 Usage Examples
Find All Offline Rooms (Efficient)
"Are any Zoom rooms offline anywhere in the company?"
→ Uses get_zoom_rooms without location_query (single API call)
Check Specific Location
"Show me all rooms in San Francisco"
→ Uses get_zoom_rooms with location_query: "SF1"
Debug Location Resolution
"How would 'DEN2' be resolved?"
→ Uses resolve_location to see what locations match
Room Status by Building
"What's the status of Denver Building 1 rooms?"
→ Uses get_zoom_rooms with location_query: "DEN1"
🏗️ Architecture
zoom-mcp/
├── src/
│ ├── server.py # Main MCP server with 5 tools
│ └── config/ # Configuration modules
│ ├── settings.py # Environment & auth configuration
│ ├── zoom_auth.py # OAuth token management
│ ├── zoom_hierarchy.py # Location discovery & relationships
│ └── zoom_fuzzy.py # Smart location resolution
├── docs/ # Comprehensive documentation
└── test_server.py # Direct testing script
Key Design Patterns
- Import Inside Functions: Configuration modules imported inside tool functions to avoid timing issues
- Multi-Level Token Caching: Memory cache + file persistence with 1-hour expiration and 5-minute buffer
- Hierarchical Discovery: Automatic campus → building → floor relationship building
- Hybrid Resolution: Hardcoded Denver aliases + dynamic fuzzy matching for other sites
🧪 Testing
MCPTools Testing
# List all tools
mcp tools uv run src/server.py
# Test authentication
mcp call test_zoom_connection --params '{}' uv run src/server.py
# Interactive testing
mcp shell uv run src/server.py
Direct Script Testing
python test_server.py
📚 Documentation
Comprehensive documentation available in docs/:
- Quick Start Guide - Setup and basic usage
- Authentication Guide - OAuth implementation details
- Testing Guide - MCPTools usage and examples
- Best Practices - Zoom-specific patterns
- Advanced Features - Deep dive into capabilities
🔒 Security
- Credentials stored in
.envfiles (not committed to git) - Token caching with secure file permissions
- Bearer token automatic refresh
- Error messages don't expose sensitive information
🤝 Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Test with MCPTools
- Submit a pull request
📄 License
This project is licensed under the MIT License.
🆘 Troubleshooting
Common Issues
-
"Zoom credentials not configured"
- Verify
.envfile exists with correct variables - Check environment variable names match exactly
- Verify
-
"Token request failed: 401"
- Verify Zoom app credentials are correct
- Ensure app has
room:read:adminscope - Confirm app is Server-to-Server OAuth type
-
"No location matches found"
- Check spelling of location query
- Use
get_zoom_sitesto see available locations - Test with
resolve_locationto debug fuzzy matching
-
Import timing issues
- Configuration modules imported inside tool functions
- Never import config at module level before
initialize_config()
Debug Commands
# Test connection
mcp call test_zoom_connection --params '{}' uv run src/server.py
# List all sites
mcp call get_zoom_sites --params '{}' uv run src/server.py
# Debug location resolution
mcp call resolve_location --params '{"location_query":"your_query"}' uv run src/server.py
Built with FastMCP and the Model Context Protocol.
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。