Apps Script MCP
Enables users to create, manage, and execute Google Apps Script projects through natural language. It provides comprehensive tools for code editing, function execution, deployment management, and monitoring script processes.
README
Google Automation MCP
Google Workspace APIs for AI agents — no GCP project required.
Uses clasp for authentication. No GCP console, no OAuth consent screen, no client secrets. Just authenticate and go.
Quick Start
uvx google-automation-mcp # Run server
uvx google-automation-mcp auth # Authenticate (one-time browser sign-in)
That's it. No GCP project setup. Tokens auto-refresh after initial auth.
Tip: Use the short alias
gmcpafter installing.
Why No GCP Project?
Traditional Google API setup requires:
- Create GCP project
- Enable APIs
- Configure OAuth consent screen
- Add test users
- Create OAuth credentials
- Download client_secret.json
This MCP uses clasp (Google's official Apps Script CLI) which handles OAuth without a GCP project. Same Google authentication, zero configuration.
Security: AI Never Sees Credentials
| Direct API | This MCP | |
|---|---|---|
| Credentials | AI handles tokens directly | AI never sees tokens |
| API access | Any endpoint | 50 curated tools only |
| Audit | Build your own | Every tool call logged |
The MCP acts as a security boundary. Your AI agent calls tools; the MCP handles authentication internally.
MCP Client Configuration
Claude Desktop (One-Click Install):
Download google-automation-mcp.dxt and open it. Claude Desktop will install automatically.
Claude Code (~/.mcp.json):
{
"mcpServers": {
"google": {
"type": "stdio",
"command": "uvx",
"args": ["google-automation-mcp"]
}
}
}
Claude Desktop (Manual) (claude_desktop_config.json):
{
"mcpServers": {
"google": {
"command": "uvx",
"args": ["google-automation-mcp"]
}
}
}
Gemini CLI:
gemini extensions install github:sam-ent/google-automation-mcp
Available Tools (50)
Gmail (5)
search_gmail_messages · get_gmail_message · send_gmail_message · list_gmail_labels · modify_gmail_labels
Drive (10)
search_drive_files · list_drive_items · get_drive_file_content · create_drive_file · create_drive_folder · delete_drive_file · trash_drive_file · share_drive_file · list_drive_permissions · remove_drive_permission
Sheets (6)
list_spreadsheets · get_sheet_values · update_sheet_values · append_sheet_values · create_spreadsheet · get_spreadsheet_metadata
Calendar (5)
list_calendars · get_events · create_event · update_event · delete_event
Docs (5)
get_doc_content · search_docs · create_doc · modify_doc_text · append_doc_text
Apps Script (17)
list_script_projects · get_script_project · get_script_content · create_script_project · update_script_content · delete_script_project · run_script_function · create_deployment · list_deployments · update_deployment · delete_deployment · list_versions · create_version · get_version · list_script_processes · get_script_metrics · generate_trigger_code
Auth (2)
start_google_auth · complete_google_auth
Multi-User Support
All tools accept user_google_email for per-user credential isolation:
search_gmail_messages(user_google_email="alice@example.com", query="is:unread")
search_gmail_messages(user_google_email="bob@example.com", query="is:unread")
Credentials stored separately: ~/.secrets/google-automation-mcp/credentials/{email}.json
Apps Script: Extending Google Workspace
Apps Script tools let you deploy code that runs inside Google apps — things REST APIs cannot do:
| Capability | Example |
|---|---|
| Custom spreadsheet functions | =VALIDATE_EMAIL(A1) in cells |
| Real-time triggers | onEdit, onOpen |
| Custom menus | Add menu items to Sheets/Docs |
| Webhooks | doGet/doPost handlers |
# Create a bound script with custom function
create_script_project(title="Validator", parent_id="SPREADSHEET_ID")
update_script_content(script_id="...", files=[{
"name": "Code",
"type": "SERVER_JS",
"source": "function VALIDATE_EMAIL(e) { return /^[^@]+@[^@]+\\.[^@]+$/.test(e); }"
}])
Limitations
run_script_function requires one-time setup per script: Open script at script.google.com → Project Settings → Change GCP project → Deploy as API Executable. Once configured, functions can be called repeatedly. All other tools work without this setup.
API quotas: Google enforces rate limits.
Production: OAuth 2.1
For multi-user deployments requiring your own OAuth credentials:
export GOOGLE_OAUTH_CLIENT_ID='...'
export GOOGLE_OAUTH_CLIENT_SECRET='...'
gmcp auth --oauth21
CLI Reference
Short alias: gmcp (or full name: google-automation-mcp)
gmcp # Run server
gmcp setup # Interactive setup wizard
gmcp auth # Authenticate with clasp
gmcp auth --oauth21 # OAuth 2.1 for production
gmcp status # Check auth status
gmcp version # Show version
Development
git clone https://github.com/sam-ent/google-automation-mcp.git
cd google-automation-mcp
uv sync
uv run pytest tests/ -v # 45 tests
Acknowledgments
Built on google_workspace_mcp by Taylor Wilsdon (MIT License).
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.
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.
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.
E2B
Using MCP to run code via e2b.