About platform-ai
platform-ai is the Google Maps Platform Code Assist toolkit, an MCP server that grounds LLM answers in official Google Maps Platform documentation and code samples. It is aimed at developers building map-based applications, and the README notes that the npm version is deprecated in favour of the Google-hosted remote server.
Key features
- 01Searches official Google Maps Platform docs and samples
- 02Optional filter by API or product area
- 03Foundational Google Maps Platform best-practice instructions
- 04Gemini CLI extension installation option
- 05Connection to Google-hosted remote MCP service
Use cases
- 01Checking official samples while integrating the Routes API
- 02Grounding a coding agent in current mapping platform docs
- 03Comparing Geocoding and Address Validation for a given use case
MCP support
The project is an MCP server exposing the retrieve-google-maps-platform-docs and retrieve-instructions tools, and it is also available as a remote endpoint at mapscodeassist.googleapis.com/mcp.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
2Saved tool list · updated 2w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| prompt | string | Yes | You are an expert prompt engineer for a Retrieval-Augmented Generation (RAG) system. **Instructions:** 1. Analyze the user's intent (e.g., are they trying to implement, troubleshoot, or learn?). 2. Identify the Google Maps Platform product and feature the user is asking about. 3. You must keep all details provided by the user in the original query. 4. Do not remove key information provided in the request, such as city, country, address, or lat/lng. 5. Add extra information that is relevant to the RAG system without removing user provided information. |
| search_context | array | No | Supplemental context to aid the search if the prompt alone is ambiguous or too broad. Put names of existing Google Maps Platform products or features specified in the user prompt. |
Details
From the project README[][npm-pkg]
Google Maps Platform resources for AI
Description
Warning
The npm version of Google Maps Platform Code Assist is deprecated. It will no longer be available as of July 1, 2026. Please use the securely hosted remote MCP version documented in this README as the primary method of connection.
This repository contains the Google Maps Platform Code Assist toolkit, a Model Context Protocol (MCP) server that enhances the responses from large language models (LLMs) used for developing applications with the Google Maps Platform by grounding them in the official, up-to-date documentation and code samples.
🔧 Available MCP Tools (gmp-code-assist)
You have access to the gmp-code-assist MCP server with these essential tools:
retrieve-google-maps-platform-docs: Searches Google Maps Platform documentation, code samples, architecture center, and GitHub repositories via RAG.- Parameters:
llmQuery(Required string query),filter(Optional API/product area filter),source(Optional string caller identifier up to 64 chars).
- Parameters:
retrieve-instructions: Retrieves foundational context on Google Maps Platform best practices.- Parameters:
name(Required string, expected format is simply "instructions").
- Parameters:
Install the Google Maps Platform Code Assist extension for Gemini CLI
- Install the Gemini CLI (alternative installation methods)
npm install -g @google/gemini-cli
- Install the Google Maps Platform extension
-
Option 1 - Install Code Assist as a Gemini CLI extension (includes context prompt, Code Assist MCP tool, and Google Maps Platform CLI theme):
gemini extensions install https://github.com/googlemaps/platform-ai.git- Verify the installation by running
gemini mcp list.
- Verify the installation by running
-
Option 2 - Add the MCP server config manually to your
~/.gemini/settings.jsonfile to securely connect to the Google-hosted remote service:{ "mcpServers": { "gmp-code-assist": { "httpUrl": "https://mapscodeassist.googleapis.com/mcp" } } }
-
Google Maps Platform Agent Skills
To equip your AI agents with specialized, portable Google Maps Platform workflows and best practices, check out the official Google Maps Platform Agent Skills repository.
Install the agent skills in Gemini CLI:
gemini skills install https://github.com/googlemaps/agent-skills.git
Install the Google Maps Platform Code Assist toolkit for other MCP clients.
For information about installing and using the toolkit with any MCP client, as well as terms of use, see the Code Assist toolkit README.
Sample Prompts
Use these example prompts to get started with the Code Assist MCP server or supported AI agents:
- Show me how to use the Routes API and display it on a Map in React + Typescript.
- I'm upgrading my map from 2D to 3D Phtotorealistic - what are my options and how would I do that with my existing codebase?
- Change my Places API (new) Text Search API implementtaion to use Places UI Kit. Estimate the amount of cost savings per user this change will make.
- Explain the difference between Address Validation and Geocoding. Which one should I apply for my current codebase for an ecommmerce checkout use case and why?
- I need an API key - can you get one and add it to my app envrionment secrets?
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.