Esri MCP Server

An MCP (Model Context Protocol) server for querying and visualizing data from the Esri Living Atlas, with a focus on water resources and mapping.

Features

• Query various Esri Living Atlas layers including:
  • Administrative: states, counties
  • Water resources: USGS gages, rivers, dams, watersheds, impaired waters, water quality stations
  • Environmental monitoring: weather stations, RAWS stations, seismic stations, CORS stations, storm reports
• Generate interactive ArcGIS JS Maps SDK applications
• Create comprehensive water context maps with multiple layers
• Support for spatial filters and geometry queries
• Web-based frontend with AI-powered query interface

Tools

• query_layer: Query feature layers with custom filters
• query_point_layer: Query point data layers (USGS gages, water quality, weather stations, etc.)
• get_layer_fields: Get field information for layers
• get_state_geometry: Retrieve state boundaries
• query_geojson: Query layers and return GeoJSON
• save_geojson: Save GeoJSON to file
• display_geojson: Visualize GeoJSON in browser
• create_arcgis_app: Generate simple ArcGIS maps
• create_arcgis_app_with_rivers: Maps with rivers overlay
• create_water_map_context: Full water resource maps
• create_embeddable_water_map: Embeddable water maps for states

Installation

1. Clone the repo
2. Install Python dependencies: pip install fastmcp requests
3. Install Node.js dependencies for frontend: cd frontend && npm install
4. Run the MCP server: python main.py --http
5. In another terminal, run the frontend: cd frontend && npm run dev

Usage

MCP Server

Start the server: python main.py --http

Connect via MCP clients or use the provided scripts in scripts/.

Frontend

Start the frontend: cd frontend && npm run dev

Open http://localhost:5174 to access the web interface with AI-powered queries.

Command-line Client

Use clients.py for command-line interaction:

# List available tools
python clients.py list-tools

# Call a tool
python clients.py call-tool query_layer layer_name usgs-gauges where "state = 'MI'"

See scripts/ for additional example usage.

Repository Structure

• main.py: Main MCP server with Esri Living Atlas tools
• frontend/: React frontend with MCP client and AI interface
• scripts/: Test and helper scripts for various queries
• clients.py: Command-line MCP client
• .gitignore: Ignores generated files

License

MIT