gmail-mcp
Enables AI assistants to read Gmail and create drafts securely with read-only access and query whitelisting.
README
gmail-mcp
MCP server for reading Gmail and creating drafts via AI assistants (Claude, OpenClaw, etc.).
Security model:
- 🔒 Read-only by default — no send capability, ever
- 📋 Drafts only — human must open Gmail and send manually
- 🔍 Query whitelist — only configured filter queries are allowed
- 👤 Optional recipient whitelist for drafts
- 🔑 API key authentication for HTTP/SSE mode
Tools
| Tool | Description |
|---|---|
list_filters |
Show configured allowed search queries |
search_email |
Search Gmail (query must match a whitelisted filter) |
get_email |
Fetch full content of an email by ID |
create_draft |
Create a draft — NOT sent, human reviews in Gmail |
Setup
1. Google Cloud credentials
- Go to Google Cloud Console
- Create a project → Enable Gmail API
- Create OAuth2 credentials (Desktop app type)
- Download as
credentials.jsonand place in the project folder
2. Install
git clone https://github.com/mortenv/gmail-mcp
cd gmail-mcp
pip install -e .
# or with uv:
uv sync
3. Configure
cp config.example.yaml config.yaml
Edit config.yaml:
- Add your filter queries (Gmail search syntax)
- Optionally restrict draft recipients
4. Authenticate
First run opens a browser for OAuth2 login:
gmail-mcp config.yaml
Token is cached in token.json — subsequent runs don't need the browser.
5. Connect to Claude / OpenClaw
stdio mode (local):
{
"mcpServers": {
"gmail": {
"command": "gmail-mcp",
"args": ["/path/to/config.yaml"]
}
}
}
HTTP/SSE mode (remote server):
export GMAIL_MCP_API_KEY=$(python -c "import secrets; print(secrets.token_hex(32))")
gmail-mcp --http config.yaml
{
"mcpServers": {
"gmail": {
"url": "https://your-server:8001/sse",
"headers": { "X-API-Key": "your-key" }
}
}
}
Filter syntax
Uses standard Gmail search syntax:
filters:
- query: "label:inbox"
label: "Inbox"
- query: "from:boss@company.com"
label: "From boss"
- query: "label:invoices newer_than:30d"
label: "Recent invoices"
The AI can narrow results by appending terms to a base query:
"label:inbox subject:urgent" — allowed if label:inbox is whitelisted.
Security notes
credentials.jsonandtoken.jsonare gitignored — never commit them- OAuth scopes:
gmail.readonly+gmail.compose(nogmail.send) - Drafts require human review — the server has no way to send email
License
MIT
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.