Google Workspace MCP Server
A secure tool bridge to Google Docs and Gmail via OAuth 2.0, with a console approval gate for write and compose actions.
README
Google Workspace MCP Server (Python)
A FastAPI-based local server that acts as a secure tool bridge to Google Docs and Gmail via OAuth 2.0. Before performing any write or compose actions, the server displays the payload in the terminal and blocks for user confirmation.
Setup Instructions
1. Enable APIs in Google Cloud Console
- Go to the Google Cloud Console.
- Create a new project or select an existing one.
- Search for and enable the following APIs:
- Google Docs API
- Gmail API
- Configure the OAuth Consent Screen:
- Set User Type to External (or Internal if inside an organization).
- Add the scopes:
https://www.googleapis.com/auth/documentshttps://www.googleapis.com/auth/gmail.compose
- Add your test email address as a Test User (required while in testing status).
- Create credentials:
- Go to the Credentials page.
- Click Create Credentials -> OAuth Client ID.
- Select application type: Desktop App.
- Click Create, then download the client secrets JSON file.
- Save the downloaded file inside this directory as
credentials.json.
2. Install Dependencies
Create a virtual environment and install the required Python packages:
# Windows
python -m venv venv
venv\Scripts\activate
# Install packages
pip install -r requirements.txt
Running the Server
Start the FastAPI application:
python server.py
Authentication Flow
- On the first run, the server will detect that
token.jsonis missing and automatically launch a Google OAuth login screen in your default web browser. - Log in using your registered Google test user email.
- Click "Continue" through the safety warnings (since the app is unverified).
- Once completed, a
token.jsonfile is saved in this directory. Subsequent runs will use this token directly and bypass browser login.
API Endpoints
1. Append to Google Doc
Appends text or executes structured styling requests at the end of the document.
- URL:
POST /append_to_doc - Payload Format (Plain Text):
{
"doc_id": "YOUR_GOOGLE_DOC_ID",
"content": "Line of text to append\n"
}
- Payload Format (Structured Styles):
{
"doc_id": "YOUR_GOOGLE_DOC_ID",
"content": "[{\"insertText\":{\"text\":\"Theme Heading\\n\",\"endOfSegmentLocation\":{}}},{\"updateParagraphStyle\":{\"paragraphStyle\":{\"namedStyleType\":\"HEADING_3\"},\"fields\":\"namedStyleType\",\"range\":{\"startIndex\":1,\"endIndex\":14}}}]"
}
2. Create Email Draft
Creates a draft inside your Gmail account.
- URL:
POST /create_email_draft - Payload Format:
{
"to": ["stakeholder1@example.com", "stakeholder2@example.com"],
"subject": "Groww Weekly Pulse — 2026-W24",
"body": "Optional plain text body",
"html_body": "<div><h2>HTML Teaser</h2><p>Teaser details...</p></div>",
"text_body": "Plain text teaser details..."
}
Console Approval Gate
When a POST request is received, the server blocks execution and displays an approval request in the terminal:
==================================================
ACTION REQUESTED: append_to_doc
PAYLOAD:
{
"doc_id": "YOUR_GOOGLE_DOC_ID",
"content": "Line of text to append\n"
}
==================================================
Approve? (y/n):
Only typing y or yes will execute the Google API request. Any other input returns an HTTP 403 Forbidden error to the requester.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。