Gmail MCP Server
Enables AI assistants to send, search, and read emails via Gmail using OAuth2 authentication.
README
Gmail MCP Server 📧
A powerful Model Context Protocol (MCP) server that enables AI assistants like Claude to interact with Gmail. This server provides seamless integration for sending emails, searching messages, and reading email content directly through Claude.
✨ Features
- Send Emails: Compose and send emails directly through Claude
- Search Emails: Use Gmail's powerful search syntax to find specific messages
- Read Emails: Retrieve and read email content by message ID
- OAuth2 Authentication: Secure authentication using Google OAuth2
- Health Check: Built-in health monitoring endpoint
🚀 Quick Start
Prerequisites
- Python 3.10 or higher
- A Google Cloud Platform account
- Claude Desktop App or any MCP-compatible client
Installation
-
Clone the repository
git clone https://github.com/Arpit-saxena-2004/gmail_mcp_server.git cd gmail_mcp_server -
Create a virtual environment
python -m venv .venv # On Windows .venv\Scripts\activate # On macOS/Linux source .venv/bin/activate -
Install dependencies
pip install -r requirements.txt
🔑 Setting Up Google OAuth2 Credentials
Step 1: Create a Google Cloud Project
- Go to Google Cloud Console
- Create a new project or select an existing one
- Enable the Gmail API for your project:
- Navigate to "APIs & Services" > "Library"
- Search for "Gmail API"
- Click "Enable"
Step 2: Configure OAuth Consent Screen
- Go to "APIs & Services" > "OAuth consent screen"
- Choose "External" user type (or "Internal" if you have a Google Workspace)
- Fill in the required information:
- App name
- User support email
- Developer contact information
- Add scopes:
https://www.googleapis.com/auth/gmail.modify - Add test users (your email address)
Step 3: Create OAuth2 Credentials
- Go to "APIs & Services" > "Credentials"
- Click "Create Credentials" > "OAuth client ID"
- Choose "Desktop app" as the application type
- Name it (e.g., "Gmail MCP Server")
- Click "Create"
- Download the JSON file
- Rename it to
credentials.jsonand place it in the project root directory
⚙️ Configuration
Connecting to Claude Desktop
-
Locate your Claude Desktop configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
-
Add the MCP server configuration:
{ "mcpServers": { "gmail": { "command": "python", "args": [ "C:\\path\\to\\your\\gmail_mcp_server\\main.py" ] } } }Note: Replace
C:\\path\\to\\your\\gmail_mcp_server\\main.pywith the actual path to yourmain.pyfile. -
Restart Claude Desktop
First-Time Authentication
- When you first use the server, it will open a browser window
- Sign in with your Google account
- Grant the requested permissions
- The server will save a
token.jsonfile for future use - You won't need to authenticate again unless you revoke access
📖 Usage Examples
Once connected to Claude, you can use natural language to interact with Gmail:
Sending Emails
"Send an email to john@example.com with subject 'Meeting Tomorrow' and tell him about the 2pm meeting"
Searching Emails
"Search my emails for messages from Sarah in the last week"
Reading Emails
"What's the latest email in my inbox?"
🛠️ Available Tools
health_check()
Check if the Gmail MCP server is running properly.
send_email(to: str, subject: str, body: str)
Send an email via Gmail.
- to: Recipient email address
- subject: Email subject line
- body: Email content
search_emails(query: str, max_results: int = 5)
Search emails using Gmail's search syntax.
- query: Gmail search query (e.g., "from:john@example.com", "subject:meeting")
- max_results: Maximum number of results to return (default: 5)
read_email(message_id: str)
Read a specific email by its message ID.
- message_id: The unique identifier of the email
🔒 Security Notes
- Never commit
credentials.jsonortoken.jsonto version control - These files are already included in
.gitignore - Store credentials securely and rotate them periodically
- Only grant necessary scopes (currently:
gmail.modify)
📁 Project Structure
gmail_mcp_server/
├── main.py # Main MCP server implementation
├── credentials.json # OAuth2 credentials (not in repo)
├── token.json # OAuth2 token (auto-generated)
├── requirements.txt # Python dependencies
├── pyproject.toml # Project metadata
├── .gitignore # Git ignore rules
└── README.md # This file
🐛 Troubleshooting
"credentials.json not found"
- Make sure you've downloaded the OAuth2 credentials from Google Cloud Console
- Rename the file to exactly
credentials.json - Place it in the project root directory
Authentication window doesn't open
- Check if port 0 (random available port) is not blocked by firewall
- Try running the server manually first:
python main.py
"Invalid credentials" error
- Delete
token.jsonand authenticate again - Verify your OAuth2 consent screen is properly configured
- Ensure you've added your email as a test user
Claude can't connect to the server
- Verify the path in
claude_desktop_config.jsonis correct - Make sure you're using absolute paths
- Restart Claude Desktop after configuration changes
🤝 Contributing
Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.
- Fork the repository
- Create your feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
📝 License
This project is licensed under the MIT License - see the LICENSE file for details.
🙏 Acknowledgments
- Built with FastMCP - A fast, simple framework for building MCP servers
- Uses Google Gmail API
- Designed for Claude by Anthropic
📞 Support
If you encounter any issues or have questions:
- Open an issue on GitHub
- Check existing issues for solutions
- Review the MCP documentation
🌟 Star History
If you find this project useful, please consider giving it a star! ⭐
Made with ❤️ for the Claude community
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。