py-mcp-google-toolbox
An MCP server that provides AI assistants with powerful tools to interact with Google services, including Gmail, Google Calendar, Google Drive, and Google Search.
Overview
py-mcp-google-toolbox provides the following Google-related functionalities:
- Gmail operations (read, search, send, modify)
- Google Calendar management (events creation, listing, updating, deletion)
- Google Drive interactions (search, read files)
- Google Search integration (search web)
Table of Contents
- Prerequisites
- Installation
- Configure MCP Settings
- Tools Documentation
- Gmail Tools
- Calendar Tools
- Drive Tools
- Search Tools
- Development
- License
Prerequisites
- Python: Install Python 3.12 or higher
- Google Cloud Console Setup:
- Go to Google Cloud Console
- Create a new project or select an existing one
- Enable the Service API:
- Go to “APIs & Services” > “Library”
- Search for and enable “Gmail API”
- Search for and enable “Google Calendar API”
- Search for and enable “Google Drive API”
- Search formand enable “Custom Search API”
- Set up OAuth 2.0 credentials from GCP:
- Go to “APIs & Services” > “Credentials”
- Click “Create Credentials” > “OAuth client ID”
- Choose “Web application”
- Note down the Client ID and Client Secret
- Client ID
- Client Secret
- download secret json and rename to credentials.json
- Generate an API key
- Go to Custom Search Engine and get its ID
Installation
Git Clone
git clone https://github.com/jikime/py-mcp-google-toolbox.git
cd py-mcp-google-toolbox
Configuration
- Install UV package manager:
curl -LsSf https://astral.sh/uv/install.sh | sh
- Create and activate virtual environment:
uv venv -p 3.12
source .venv/bin/activate # On MacOS/Linux
# or
.venvScriptsactivate # On Windows
- Install dependencies:
uv pip install -r requirements.txt
- Get refresh token (if token is expired, you can run this)
uv run get_refresh_token.py
This will:
- Open your browser for Google OAuth authentication
- Request the following permissions:
https://www.googleapis.com/auth/gmail.modifyhttps://www.googleapis.com/auth/calendarhttps://www.googleapis.com/auth/gmail.sendhttps://www.googleapis.com/auth/gmail.readonlyhttps://www.googleapis.com/auth/drivehttps://www.googleapis.com/auth/drive.filehttps://www.googleapis.com/auth/drive.readonly
- Save the credentials to
token.json - Display the refresh token in the console
- Environment variables:
cp env.example .env
vi .env
# change with your key
GOOGLE_API_KEY=your_google_api_key
GOOGLE_CSE_ID=your_custom_search_engine_id
GOOGLE_CLIENT_ID=your_google_client_id
GOOGLE_CLIENT_SECRET=your_google_client_secret
GOOGLE_REFRESH_TOKEN=your_google_refresh_token
- copy credentials.json to project root folder (py-mcp-google-toolbox)
Using Docker
- Build the Docker image:
docker build -t py-mcp-google-toolbox .
- Run the container:
docker run py-mcp-google-toolbox
Using Local
- Run the server:
mcp run server.py
- Run the MCP Inspector
mcp dev server.py
Configure MCP Settings
Add the server configuration to your MCP settings file:
Claude desktop app
- To install automatically via Smithery:
npx -y @smithery/cli install @jikime/py-mcp-google-toolbox --client claude
- To install manually
open
~/Library/Application Support/Claude/claude_desktop_config.json
Add this to the mcpServers object:
{
"mcpServers": {
"Google Toolbox": {
"command": "/path/to/bin/uv",
"args": [
"--directory",
"/path/to/py-mcp-google-toolbox",
"run",
"server.py"
]
}
}
}
Cursor IDE
open ~/.cursor/mcp.json
Add this to the mcpServers object:
{
"mcpServers": {
"Google Toolbox": {
"command": "/path/to/bin/uv",
"args": [
"--directory",
"/path/to/py-mcp-google-toolbox",
"run",
"server.py"
]
}
}
}
for Docker
{
"mcpServers": {
"Google Toolbox": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"py-mcp-google-toolbox"
]
}
}
}
Tools Documentation
Gmail Tools
list_emails: Lists recent emails from Gmail inbox with filtering optionssearch_emails: Performs advanced Gmail searches with detailed email content retrievalsend_email: Composes and sends emails with support for CC, BCC recipientsmodify_email: Changes email states (read/unread, archived, trashed) by modifying labels
Calendar Tools
list_events: Retrieves upcoming calendar events within specified time rangescreate_event: Creates new calendar events with attendees, location, and descriptionupdate_event: Modifies existing calendar events with flexible parameter updatingdelete_event: Removes calendar events by event ID
Drive Tools
read_gdrive_file: Reads and retrieves content from Google Drive filessearch_gdrive: Searches Google Drive for files with customizable queries
Search Tools
search_google: Performs Google searches and returns formatted results
Development
For local testing, you can use the included client script:
# Example: List emails
uv run client.py list_emails max_results=5 query="is:unread"
# Example: Search emails
uv run client.py search_emails query="from:test@example.com"
# Example: Send email
uv run client.py send_email to="test@example.com" subject="test mail" body="Hello"
# Example: Modify email
uv run client.py modify_email id=MESSAGE_ID remove_labels=INBOX add_labels=ARCHIVED
# Example: List events
uv run client.py list_events time_min=2025-05-01T00:00:00+09:00 time_max=2025-05-02T23:59:59+09:00 max_results=5
# Example: Create event
uv run client.py create_event summary="new event" start=2025-05-02T10:00:00+09:00 end=2025-05-02T11:00:00+09:00 attendees="user1@example.com,user2@example.com"
# Example: Update event
uv run client.py update_event event_id=EVENT_ID summary="update event" start=2025-05-02T10:00:00+09:00 end=2025-05-02T11:00:00+09:00 attendees="user1@example.com,user2@example.com"
# Example Delete event
uv run client.py delete_event event_id=EVENT_ID
# Example: Search Google
uv run client.py search_google query="what is the MCP?"
# Example: Search Google Drive
uv run client.py search_gdrive query=mcp
# Example: Read file
uv run client.py read_gdrive_file file_id=1234567890
License
MIT License
Google Toolbox
Project Details
- jikime/py-mcp-google-toolbox
- Last Updated: 5/12/2025
Recomended MCP Servers
The Cursor Memory MCP Server provides a powerful tool to assist in coding with cursor, enpower cursor to...
Postgres MCP server with configurable auth
Connect your Sanity content to AI agents. Create, update, and explore structured content using Claude, Cursor, and VS...
加密mcp服务器,crypto mcp
A Model Context Protocol (MCP) server that helps AI code editors find TypeScript symbol definitions in your codebase....
MCP server implementation for Telegram
Model Context Protocol (MCP) Server for National Park Services data
MCP for get a UUID
test server for semantic tool search
A programming framework for agentic AI 🤖 PyPi: autogen-agentchat Discord: https://aka.ms/autogen-discord Office Hour: https://aka.ms/autogen-officehour
(*^__^*) 嘻嘻……





