mcp-gmail
Enables interaction with Gmail through the Gmail API to read unread messages, retrieve email threads, and create draft replies programmatically. Supports OAuth authentication and automated email processing through natural language.
README
mcp-gmail
A toy demo of an MCP server that can write you some nifty draft emails
https://github.com/user-attachments/assets/2dcba8aa-81c5-4098-bccd-d07a6324be7c
Requirements
- uv
- claude desktop
- a google workspace setup with desktop application access for the Gmail API (steps Setup Your Environment through Authorize Credentials For a Desktop Application) with
gmail.readonlyandgmail.composescopes
Example claude_desktop_config.json
{
"mcpServers": {
"email": {
"command": "uv",
"args": ["--directory", "/your/path/to/mcp-gmail", "run", "main.py"]
},
"preferences": {
"quickEntryShortcut": "off"
}
}
Getting started in Claude Desktop
- Clone the repo,
- Add the MCP server to your
claude_desktop_config.json(see above) - Quit and restart the application
- The MCP server should run via the
stdiotransport locally; you can check it's enabled in the 'connectors' section of the prompt (click the+icon)
What The Server Does
This repository implements a Gmail automation server using FastMCP and the Google Gmail API. The gmail.py file contains a Gmail class that handles authentication with the Gmail API and provides methods for interacting with Gmail, such as get_unread_messages to retrieve unread emails, get_thread to fetch an entire email thread, and create_draft to compose a draft reply within a specific thread. The models.py file defines Pydantic models for structuring Gmail message data, ensuring data integrity and type safety when interacting with the dynamic Gmail API responses. The main.py file sets up a FastMCP server, exposing the Gmail functionalities as callable tools: get_unread_emails, get_email_thread, and create_draft_reply. This allows external clients to interact with Gmail programmatically through a well-defined API, enabling automated email processing and drafting. The server is designed to run locally, authenticating users and performing actions on their behalf.
Notes
- I was using Claude Desktop, and my original plan was to use MCP Sampling, but Claude Desktop does not appear to support sampling; instead we just expose the tools needed to draft an email
- AuthN/AuthZ through MCP clients/servers feels somewhat complex; this just has the app deal with it when it spins up.
- The messages returned by the Gmail API are unread, but don't appear to have guaranteed ordering: the first message returned might not be the most recent message. This is likely to be undesired behaviour.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.