Provides capabilities to fetch unread or recent emails and generate draft replies based on message content.
Utilizes Google Cloud project configurations and credentials to enable secure access to the Google Docs API.
Retrieves writing style guidelines and context from specific Google Doc files to personalize and inform email drafts.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Gmail Assistant MCPfetch my last 10 emails and draft replies for the ones addressed to me"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Gmail Assistant MCP
MCP server that integrates with Gmail to fetch unread emails and generate draft emails using Claude. Also connects with a Google Doc file that outlines my writing style and useful context.
How it works
There's 3 toosl to use:
Ask Claude to fetch emails
Ask Claude to draft replies to any of those emails (where you are the "to" (not "cc"))
Ask Claude to fetch last 10 emails and draft replies for ones where they are "to" you
Requirements
Python 3.10+
Gmail account with App Password enabled
Anthropic API key
Google Cloud project with Docs API enabled (optional, for guidelines)
Installation
Clone the repo
Create an activate virtual environment
Install dependencies
Set up Gmail App Password (via https://myaccount.google.com/security)
Set up Google Docs API for guidelines (via https://console.cloud.google.com/) a. Create a project b. Enable Google Docs API c. Create credentials d. Download credentials as
credentials.jsonin project root
Configuration
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):