Google Workspace MCP Server

This is a generic Model Context Protocol (MCP) server that provides AI agents with tools to interact with Gmail and Google Docs. It allows any MCP-compatible agent to draft emails, send emails, and append content to Google Docs on your behalf.

Provided Tools

1. gmail_create_draft: Creates an email draft.

2. gmail_send_email: Sends an email immediately.

3. google_doc_append: Appends text to the end of an existing Google Document.

Related MCP server: Google Workspace MCP Server

Setup Instructions

To use this server, you need to configure a Google Cloud Project with the necessary APIs and OAuth credentials.

1. Create a Google Cloud Project

1. Go to the Google Cloud Console.

2. Create a new project.

2. Enable APIs

1. Navigate to APIs & Services > Library.

2. Search for and enable the Gmail API.

3. Search for and enable the Google Docs API.

3. Configure OAuth Consent Screen

1. Navigate to APIs & Services > OAuth consent screen.

2. Choose External (or Internal if you are a Google Workspace user).

3. Add the following scopes:

   • https://www.googleapis.com/auth/gmail.compose

   • https://www.googleapis.com/auth/gmail.send

   • https://www.googleapis.com/auth/documents

4. Add your email address as a Test user if your app is in "Testing" mode.

4. Create OAuth Credentials

1. Navigate to APIs & Services > Credentials.

2. Click Create Credentials > OAuth client ID.

3. Select Desktop app (or Web application if you prefer) as the application type.

4. Download or copy your Client ID and Client Secret.

5. Obtain a Refresh Token

You must perform an initial OAuth 2.0 flow to authorize the application and obtain a refresh token. (Note: You can use tools like Google OAuth 2.0 Playground configured with your own OAuth Client ID/Secret to generate a refresh token with the required scopes).

6. Configure Environment Variables

1. Copy the .env.example file to .env:

   cp .env.example .env

2. Fill in the .env file with your credentials:

   GOOGLE_CLIENT_ID=your_client_id
   GOOGLE_CLIENT_SECRET=your_client_secret
   GOOGLE_REFRESH_TOKEN=your_refresh_token
   MCP_SERVER_NAME=google-integration-server

7. Install and Build

npm install
npm run build

MCP Client Configuration

Configure your MCP client (such as Cursor or Claude Desktop) to connect to the compiled server.

Example for claude_desktop_config.json:

{
  "mcpServers": {
    "google-integration": {
      "command": "node",
      "args": [
        "/absolute/path/to/AI Summarizer MCP server/build/index.js"
      ],
      "env": {
        "GOOGLE_CLIENT_ID": "your_client_id",
        "GOOGLE_CLIENT_SECRET": "your_client_secret",
        "GOOGLE_REFRESH_TOKEN": "your_refresh_token",
        "MCP_SERVER_NAME": "google-integration-server"
      }
    }
  }
}

(Alternatively, rely on the .env file located in the server's root directory).

Architecture & Security

• This server relies on Google's OAuth 2.0. Access tokens are kept entirely within the server's memory and are automatically refreshed.

• Tokens and email/document contents are not exposed to the AI agent except via structured success/error messages.