Personal Google MCP Server

Personal Google MCP Server

Gives Claude read/write access to your own Gmail, Google Docs, Google Sheets, and Google Tasks via a tiny always-on web service.

Category
访问服务器

README

Personal Google MCP Server

Gives Claude read/write access to your own Gmail, Google Docs, Google Sheets, and Google Tasks. Runs as a tiny always-on web service so it works from Claude on your phone with your computer off.

Tools included: sheets_read, sheets_write, sheets_append, docs_read, docs_append_text, gmail_search, gmail_read, gmail_create_draft, gmail_send, tasks_list, tasks_create, tasks_update.

Cost: $0/month (Render free tier + Google APIs are free for personal use).


Step 1 — Create a Google Cloud project + OAuth credentials (~10 min)

  1. Go to https://console.cloud.google.com/projectcreate — create a project (any name).
  2. Go to APIs & Services → Library and enable, one by one:
    • Google Sheets API
    • Google Docs API
    • Gmail API
    • Tasks API
  3. Go to APIs & Services → OAuth consent screen.
    • User type: External. App name: anything. Add your own email as a test user.
    • You do NOT need to submit for verification — test mode works indefinitely for you as a test user.
  4. Go to APIs & Services → Credentials → Create Credentials → OAuth client ID.
    • Application type: Desktop app.
    • Save the Client ID and Client Secret shown.

Step 2 — Get a refresh token (run once, on your own computer)

You need Node.js installed locally for this one step only.

# unzip the project folder first, then:
cd google-mcp-server
npm install
node get-token.js YOUR_CLIENT_ID YOUR_CLIENT_SECRET

Open the URL it prints, approve access with your Google account, and it will print a GOOGLE_REFRESH_TOKEN=... value in your terminal. Save that — you won't do this step again unless you revoke access.

Step 3 — Deploy to Render (free, no credit card)

  1. Push this folder to a new private GitHub repo.
  2. Go to https://render.com → sign up free → New → Web Service → connect that repo.
  3. Settings:
    • Build command: npm install
    • Start command: npm start
    • Instance type: Free
  4. Add environment variables (Render dashboard → Environment):
    • GOOGLE_CLIENT_ID — from Step 1
    • GOOGLE_CLIENT_SECRET — from Step 1
    • GOOGLE_REFRESH_TOKEN — from Step 2
    • MCP_SECRET_PATH — make up a long random string (e.g. run node -e "console.log(require('crypto').randomBytes(24).toString('hex'))" and use the output). This acts as your connector's password — anyone with this exact URL can access your Google account through it, so keep it private and don't commit it to a public repo.
  5. Deploy. Render gives you a URL like https://your-app.onrender.com.

Note: Render's free tier spins the service down after 15 minutes of no traffic, so the first request after a break takes ~30–50 seconds to wake up. That's normal — just wait for it once.

Step 4 — Add it to Claude

  1. In the Claude app: Settings → Connectors → Add custom connector.
  2. Name: anything (e.g. "My Google Tools").
  3. URL: https://your-app.onrender.com/mcp/YOUR_MCP_SECRET_PATH
  4. Leave OAuth fields blank — save.
  5. In a chat, enable the connector (+ → Connectors) and try: "list my Google Tasks" or "read Sheet [ID], tab Trip".

This works identically on your phone and desktop, since the server lives on Render, not on your computer.

Notes / limits

  • This is single-user: it only ever acts as you (whoever ran Step 2). Don't share the URL.
  • gmail_send sends immediately — no confirmation step. If you'd rather always review before sending, just tell Claude to use gmail_create_draft instead, or remove the gmail_send tool from index.js.
  • To add more tools later (Calendar, Drive search, etc.), copy the pattern of any existing server.registerTool(...) block in index.js.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选