MCP Server: Google Drive

What This Repository Does

@modelcontextprotocol/server-gdrive is an MCP server that connects build-cli to Google Drive and Docs, enabling document reference directly within a session. It uses a Google Cloud service account or OAuth credentials to access Drive content.

Common use cases:

Reference a requirements doc from Google Drive while working on a technical implementation
Pull content from a Google Sheet into a build-cli session for analysis
Cross-reference a data dictionary in Drive while reviewing code

Note on tool choice: For tasks that are entirely document-centric (summarising, drafting, editing), Gemini embedded in Google Suite is the better tool. Use this MCP when you need Google Drive content alongside technical work in build-cli. See the ‘Which AI Tool to Use?’ guide for a full decision matrix.

How to Use It

Before setup: Confirm with your IT/security team that connecting a local MCP server to Google Workspace is permitted under your organisation’s policies.

Add the server to your Claude Code MCP configuration (~/.claude/settings.json):

{
  "mcpServers": {
    "gdrive": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-gdrive"],
      "env": {
        "GDRIVE_CLIENT_ID": "your-client-id",
        "GDRIVE_CLIENT_SECRET": "your-client-secret"
      }
    }
  }
}

Refer to the server’s documentation (linked in source_url) for the full OAuth setup flow.

Security considerations:

Only grant the scopes you need — read-only Drive access is sufficient for most use cases.
Do not connect to shared drives containing sensitive or confidential data without checking with your data steward.

Relevance to the Hub

Many DMDQP artefacts (business requirements, data dictionaries, STM mapping documents) live in Google Drive. This server bridges the gap between Drive-resident documents and build-cli’s technical context, enabling engineers and business analysts to work with both in a single session without manual copy-paste.