mcp-gee-sweet 0.8.2.dev224

MCP server for Google Workspace — Sheets, Drive, Docs (HTML or Markdown), and Calendar. 60+ tools for AI clients.

mcp-gee-sweet

Your AI Assistant's Gateway to Google Workspace!

An MCP server that gives AI clients reliable, direct access to Google Workspace — Sheets, Drive, Docs (HTML or Markdown content), and Calendar. 101 tools across four domains.

As of mid-2026, Google's own Workspace MCP servers (Developer Preview) cover Gmail, Drive, Calendar, People, and Chat — Sheets and Docs still have no official server at all, and Drive/Calendar coverage there is a fraction of what's here. Stable releases ship to PyPI behind a QA regression gate, with a raw batch_update passthrough as an escape hatch for anything a named tool doesn't cover yet. See Why it exists for the full, sourced comparison against both the official servers and community alternatives.

Install (stable):

uvx mcp-gee-sweet

Install (bleeding edge — every code change on develop):

uvx --prerelease=allow mcp-gee-sweet

Full documentation →

---

Quick start

Option A: uv — local use and development

If you cloned the repo or just want to use the server from your machine, uv is all you need. No Docker required.

git clone https://github.com/khuisman/mcp-gee-sweet.git
cd mcp-gee-sweet
uv sync

Then point your MCP client at it using stdio transport (see MCP client config below). The client spawns the server as a subprocess on demand — each session gets its own isolated process, so restarting the server for a code change or config update doesn't affect other open sessions.

Option B: Docker — persistent shared server (SSE)

Use Docker when you want a single long-running server that multiple clients connect to over SSE — for example, Claude Desktop talking to the same instance as Claude Code.

git clone https://github.com/khuisman/mcp-gee-sweet.git
cd mcp-gee-sweet
make build   # build the container image
make start   # start the server (SSE on port 47000)
make logs    # tail logs

Point your MCP client at http://localhost:47000/sse.

Note: when using SSE, all clients share one server process. After a code change or restart (make restart), you must also restart each MCP client to reconnect.

---

Configuration

The server tries auth methods in a waterfall by default, OAuth first — it authenticates as you and has full personal Drive access, which is what local/dev use (Option A above) needs. For OAuth, download an OAuth Client ID JSON from GCP Console and point the server at it:

export CREDENTIALS_PATH="/path/to/credentials.json"

Service accounts (recommended for headless server deployments — see Option B above), base64 credential injection, and Application Default Credentials are also supported. See Authentication for all options.

---

MCP client config

Claude Desktop — stdio (cloned repo):

{
  "mcpServers": {
    "mcp-gee-sweet": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/mcp-gee-sweet", "mcp-gee-sweet"],
      "env": {
        "CREDENTIALS_PATH": "/path/to/credentials.json"
      }
    }
  }
}

Claude Desktop — SSE (Docker):

{
  "mcpServers": {
    "mcp-gee-sweet": {
      "transport": "sse",
      "url": "http://localhost:47000/sse"
    }
  }
}

See Client Setup for more options including tool filtering.

---

Docs

• Tools — full tool reference (101 tools)
• Authentication — all four auth methods
• Configuration — env vars, caching, tool filtering
• Client Setup — MCP client config examples
• Design Principles — tool inclusion policy, composite tool decisions
• Roadmap — planned features and known gaps

---

Contributing

See CONTRIBUTING.md for setup, QA workflows, and PR guidelines. A signed-off docs/qa/runs/vX.Y.Z.md record is required before any stable PyPI release — see docs/qa/README.md for the QA workflow.

License

MIT — see LICENSE.

Credits

• xing5/mcp-google-sheets — upstream this project was forked from
• FastMCP
• freema/mcp-gsheets and piotr-agier/google-drive-mcp — roadmap inspiration