AI
MCP servers and agent skills that help AI assistants work with MapLibre: check and render styles, write better map code, and drive a live map.
MCP servers
Model Context Protocol servers give an AI assistant tools of its own: checking and rendering styles, or driving a live map.
MapControl
A live, shareable map that AI agents drive
An MCP server, Python SDK, and REST API from ESIP for a live MapLibre map: agents create maps, add and style GeoJSON and GeoTIFF data, switch basemaps, move the camera, turn on 3D terrain and the globe, and take screenshots, while people watch the same map update in their browser and share it by link.
maplibre-mcp
Check, render, and show MapLibre styles from AI agents
An MCP server that gives AI assistants MapLibre tools: validate a style against the MapLibre Style Specification, check its layers against the data in its sources, render it with MapLibre GL JS, MapLibre Native, or a Martin server, compare two versions, show the user an interactive map, and search this directory for libraries and basemaps. It runs locally with npx and needs no API key.
map-mcp
Let an AI agent read and drive a live MapLibre map
An MCP server and command line tool that reach into a MapLibre GL JS map already on a web page: an agent can query the rendered features, read the viewport, click and read popups, navigate, and toggle layers on the same map a person is looking at. The page adds a one-line hook.
Agent skills
Guidance that coding assistants read before they write MapLibre code, so they avoid its common mistakes.
This directory is itself available to agents: see Using the data, or search it from an assistant with maplibre-mcp.
Know of another AI tool for MapLibre?
Suggest MCP servers, agent skills, and other AI tools for MapLibre.