Skip to content
MCP Server Directory logo
developer-tools

Google Maps Code Assist

Reviewed overview

MCP サーバー概要

What is Code Assist? Code Assist (Google Maps Platform Code Assist) is an MCP (Model Context Protocol) server hosted by Google that enhances LLM/AI agent responses for Google Maps Platform development by grounding answers in official, up-to-date documentation, code samples, Terms of Service, Trust Center content, and GitHub repositories via retrieval-augmented generation (RAG).

How to use it?

  • Connect your AI agent to the remote streamable HTTP MCP endpoint: https://mapscodeassist.googleapis.com/mcp
  • Add the server to your agent’s MCP configuration (examples provided for Gemini CLI, Claude Code, Cursor, Codex, Antigravity, Android Studio). No authentication is required for the hosted endpoint.
  • Note: the NPM version is being deprecated; use the remote streamable HTTP endpoint instead.

Key features

  • Grounds LLM responses in official Maps Platform docs, code samples, architecture guides, and repo content (RAG).
  • Exposes tools for AI clients: retrieve-instructions (best-practice context) and retrieve-google-maps-platform-docs (document/code search by llmQuery, optional filters and source).
  • Works with any agent that supports MCP (Gemini, Claude Code, Codex, Cursor, Antigravity, etc.).
  • Streamable HTTP transport with a /mcp POST endpoint per the MCP spec.
  • Open-source repository with license and GitHub issue/contribution support.

Primary use cases

  1. Producing accurate, up-to-date Maps Platform sample code and snippets.
  2. Troubleshooting/integration guidance for Maps APIs (Maps JS API, Places UI Kit, 3D features, etc.).
  3. Ensuring compliance and referencing Terms of Service or Trust Center items in generated guidance.
  4. Accelerating developer workflows by augmenting AI agents with authoritative Maps Platform context.

FAQ

  • Is authentication required to use the hosted MCP endpoint?

No — the remote streamable HTTP endpoint is publicly accessible per the toolkit docs.

  • Where does the server source its content?

It searches Google Maps Platform documentation, official GitHub org repositories, architecture center, Terms of Service, and Trust Center.

  • Will this work with my AI agent?

Yes, any agent that supports MCP streamable HTTP transports can be configured to use the server (examples provided for major agents).

  • Is the project open source and supported by Google?

The toolkit is provided under an open-source license in the repo. It is hosted by Google but is not a Maps Platform Core Service; platform services remain subject to Google Maps Platform Terms of Service.

  • What about local installation or npm packages?

The NPM version is being deprecated; the recommended approach is to use the hosted streamable HTTP endpoint.