Skip to main content
MCP server CX Codex

ArcGIS MCP Bridge — Claude Code MCP server: setup

This MCP server exposes ArcPy and ArcGIS Pro tools to coding agents. Use it to run GIS operations from natural-language requests.

Install into Claude Code · Platform CLI All methods and platforms ↓
claude mcp add --env ARCPY_PYTHON_PATH=<ARCPY_PYTHON_PATH> --env ARCGIS_MCP_ALLOWED_ROOTS=<ARCGIS_MCP_ALLOWED_ROOTS> --env ARCGIS_MCP_SCRATCH_GDB=<ARCGIS_MCP_SCRATCH_GDB> --env ARCGIS_MCP_MAX_WORKERS=<ARCGIS_MCP_MAX_WORKERS> --transport stdio arcgis-mcp-bridge -- uvx arcgis-mcp-bridge
Open on GitHub

What ArcGIS MCP Bridge does

ArcGIS MCP Bridge gives an agent access to a catalog of 100 ArcGIS Pro operations, including geodatabase work, layer handling, spatial analysis, geometry checks, and map tasks. Communication uses JSON-RPC over stdio, while failures are returned as structured results for validation, paths, licensing, and geoprocessing.

The bridge keeps the MCP-facing server separate from the ArcPy worker. The first process handles requests and starts jobs; the worker runs on the ArcGIS Pro Python interpreter. PathGuard limits filesystem access to approved roots. Three core endpoints are included: health_check, list_layers, and execute_spatial_tool.

Windows, ArcGIS Pro, and Python 3.11 or newer are required. Set ARCPY_PYTHON_PATH to the ArcGIS Pro environment that contains the package. A different interpreter can make the worker fail when it loads ArcPy or native extensions.

What it helps with

  • run ArcGIS Pro geoprocessing from an agent
  • inspect feature classes inside a geodatabase
  • create buffers and spatial analysis outputs
  • check and repair geometry issues
  • calculate terrain values from raster data

Example prompts for the agent

  • List feature classes in C:\GIS\city.gdb whose names start with road_.
  • Buffer every parcel in my geodatabase by 50 meters and save the result to scratch.
  • Calculate slope and aspect from C:\GIS\dem.tif.

How to install ArcGIS MCP Bridge

Platform CLI

claude mcp add --env ARCPY_PYTHON_PATH=<ARCPY_PYTHON_PATH> --env ARCGIS_MCP_ALLOWED_ROOTS=<ARCGIS_MCP_ALLOWED_ROOTS> --env ARCGIS_MCP_SCRATCH_GDB=<ARCGIS_MCP_SCRATCH_GDB> --env ARCGIS_MCP_MAX_WORKERS=<ARCGIS_MCP_MAX_WORKERS> --transport stdio arcgis-mcp-bridge -- uvx arcgis-mcp-bridge

Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from arcgis-mcp-bridge <command>

Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more

Questions about ArcGIS MCP Bridge

How do I install ArcGIS MCP Bridge in Claude Code?

Run claude mcp add arcgis-mcp-bridge -e ARCPY_PYTHON_PATH=<ARCPY_PYTHON_PATH> -- uvx arcgis-mcp-bridge.

What does ArcGIS MCP Bridge do?

It connects ArcPy and ArcGIS Pro operations to an MCP host. An agent can work with layers, geodatabases, and geoprocessing tasks through plain-language requests.

What does the server require?

It needs Windows, a licensed ArcGIS Pro setup, and Python 3.11 or newer. ARCPY_PYTHON_PATH must point to the ArcGIS Pro Python environment.

Similar MCP servers

All MCP servers for Agent setup: memory, context, orchestration →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor