gmail-cleanup-mcp

gmail-cleanup-mcp

Enables bulk Gmail inbox cleanup including trashing emails, managing filters, and extracting unsubscribe links, with dry-run previews and automatic retry/backoff to handle rate limits.

Category
Visit Server

README

Gmail Cleanup MCP Server

A focused Model Context Protocol server for Gmail inbox cleanup — trash emails, manage filters, and extract unsubscribe links. Designed to complement Anthropic's built-in Gmail connector, which handles read, search, draft, and send.

Patched fork of dcerniglia/gmail-cleanup-mcp (MIT licensed). This fork fixes a bug where search_and_trash and batch_trash_emails silently dropped failures instead of retrying — firing 50 concurrent trash requests could trip Gmail's rate limit and drop ~half a batch with no error shown. Now both tools batch smaller, back off, auto-retry, and report the real error per failed message.

Tools

Tool Description
trash_email Move a single email to trash (30-day safety net)
batch_trash_emails Trash up to 500 emails by message ID, processed in batches of 50
search_and_trash Find emails by Gmail query and trash them — supports dry-run preview
create_filter Create filters to auto-archive, auto-label, auto-delete, or forward
list_filters List all existing Gmail filters with criteria and actions
delete_filter Remove a Gmail filter by ID
get_unsubscribe_info Extract List-Unsubscribe headers from an email

Prerequisites

  • Node.js 18+
  • A Google Cloud project with the Gmail API enabled
  • OAuth 2.0 credentials (Desktop app type)

Setup

1. Google Cloud OAuth

Already have ~/.gmail-mcp/gcp-oauth.keys.json from another Gmail MCP server? Skip to step 2 — this server uses the same credential location.

  1. Go to Google Cloud Console
  2. Create a project (or reuse an existing one) and enable the Gmail API
  3. Go to CredentialsCreate CredentialsOAuth client ID
  4. Choose Desktop app, download the JSON
  5. Save it as ~/.gmail-mcp/gcp-oauth.keys.json (or place gcp-oauth.keys.json in the project directory — it will be copied automatically on first run)

2. Install & Build

git clone https://github.com/davidcerniglia/gmail-cleanup-mcp.git
cd gmail-cleanup-mcp
npm install
npm run build

3. Authenticate

npm run auth

This opens a browser for Google OAuth consent. Credentials are saved to ~/.gmail-mcp/credentials.json with 0600 permissions (owner-only).

4. Add to Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "gmail-cleanup": {
      "command": "node",
      "args": ["/absolute/path/to/gmail-cleanup-mcp/dist/index.js"]
    }
  }
}

Or for Claude Code, add to .claude/settings.json:

{
  "mcpServers": {
    "gmail-cleanup": {
      "command": "node",
      "args": ["/absolute/path/to/gmail-cleanup-mcp/dist/index.js"]
    }
  }
}

Usage Examples

Prompt What happens
"Trash all emails from newsletter@spam.com older than 6 months" search_and_trash with Gmail query syntax
"Do a dry run — show me what would get trashed from LinkedIn" search_and_trash with dryRun: true previews matches first
"Create a filter to auto-archive emails from nextdoor.com" create_filter with removeLabelIds: ["INBOX"]
"Check if that marketing email has an unsubscribe link" get_unsubscribe_info extracts the List-Unsubscribe header

How it works with Claude's built-in Gmail

Built-in Gmail connector This server
Search & read Yes
Draft & send Yes
Trash emails Yes
Bulk cleanup Yes (up to 500/call)
Filter management Yes
Unsubscribe extraction Yes

Security

  • Minimal scopes: gmail.modify + gmail.settings.basic — the minimum needed for trash and filter operations
  • Trash, not delete: All deletions use Gmail's trash (30-day recovery window), never permanent delete
  • Dry-run mode: search_and_trash can preview matches before committing
  • Restrictive file permissions: Credentials stored with 0600 (owner read/write only)
  • No attachment handling: Zero path traversal risk
  • No filesystem access beyond credential storage in ~/.gmail-mcp/

Environment Variables

Variable Default Description
GMAIL_OAUTH_PATH ~/.gmail-mcp/gcp-oauth.keys.json Path to OAuth client keys
GMAIL_CREDENTIALS_PATH ~/.gmail-mcp/credentials.json Path to stored credentials

License

MIT

Recommended Servers

playwright-mcp

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.

Official
Featured
TypeScript
Audiense Insights MCP Server

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.

Official
Featured
Local
TypeScript
Magic Component Platform (MCP)

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.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

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.

Official
Featured
TypeScript
Kagi MCP Server

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.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

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.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured