
Google Docs
Let your agent read and edit Google Docs for specs, runbooks, and launch copy while you stay in the coding workflow.
Overview
Google Documents MCP is a Build-phase MCP server that enables AI agents to read and edit Google Docs via stdio or streamable HTTP.
What is this MCP server?
- Read and edit Google Docs from MCP-compatible agents
- Version 1.2.0 npm package with stdio or streamable-http transports
- OAuth: GOOGLE_ACCESS_TOKEN with Docs scope (stdio) or client credentials for HTTP mode
- npx runtimeHint for google-documents-mcp
- Supports PRDs, changelogs, and shared specs without manual copy-paste from the browser
- Package version 1.2.0
- Registry identifier google-documents-mcp on npm
- 2 package variants: stdio and streamable-http
Community signal: 1 GitHub stars.
What problem does it solve?
Your spec and doc truth lives in Google Docs but your agent only sees the repo, forcing constant manual export or paste.
Who is it for?
Indie builders who document in Google Docs and want the agent to sync specs, update sections, or ingest doc context during implementation.
Skip if: Workflows that need full Drive browsing, binary uploads, or offline-first doc editing without Google APIs.
What do I get? / Deliverables
Your agent can open Docs content and apply edits through MCP after OAuth is configured.
- MCP tools to read and edit Google Docs documents
- v1.2.0 npm-deployed server with dual transport configuration
Recommended MCP Servers
Journey fit
Documentation artifacts are created and revised continuously during Build; this MCP is the Google Docs bridge for that work. Docs subphase is the canonical shelf because the server’s purpose is read/edit Google Docs, not generic file storage or contacts.
How it compares
Docs-focused Google MCP server, not Google Drive file management or a markdown-only docs skill.
Common Questions / FAQ
Who is Google Documents MCP for?
Solo builders and small teams whose living documents are in Google Docs and who use MCP agents to read or mutate that content during development.
When should I use Google Documents MCP?
Use it when drafting or updating PRDs, runbooks, or shared notes in Docs while the agent implements or reviews code in the same session.
How do I add Google Documents MCP to my agent?
Register google-documents-mcp with npx in MCP settings, supply a Google OAuth token with Docs scope for stdio, or run HTTP transport with MCP_TRANSPORT=http and OAuth client env vars.