Open Streetmap MCP

What jagan-shanmugam/open-streetmap-mcp MCP server does

The jagan-shanmugam/open-streetmap-mcp MCP server makes OpenStreetMap data available to MCP-compatible clients through location tools and resources. It covers both basic geographic operations and higher-level queries about places, neighborhoods, travel, and infrastructure.

Core use cases include turning an address or place name into coordinates, converting coordinates into a readable address, finding points of interest near a location, and obtaining turn-by-turn directions between locations. It also supports category-based place searches within an area, such as finding locations that match a specified type.

The server includes tools for more specialized decisions. Agents can compare commute options between home and work, identify schools, filter EV charging stations by connector or power information, locate parking facilities with availability and fee details, suggest meeting points for several people, and assess neighborhood livability for real estate use cases.

How it works

The jagan-shanmugam/open-streetmap-mcp MCP server communicates with clients over MCP using standard input and output when launched locally. MCP tools accept structured arguments and return location-related results for an agent to use in a conversation or workflow.

It also exposes two location-oriented resource patterns. The location://place/{query} resource provides information about a place identified by a name or address. The location://map/{style}/{z}/{x}/{y} resource provides styled map tiles for specified map coordinates and zoom values.

The repository includes example Python clients. One demonstrates connecting to a locally running server, retrieving information about San Francisco, searching for restaurants, and obtaining map data with progress tracking. Another example shows an LLM-oriented location assistant for text-based location queries, nearby places, directions, meeting points, and neighborhood exploration.

Setup and configuration

The published server can be configured in an MCP host with the osm-mcp-server package and the uvx runner:

{
  "mcpServers": {
    "osm-mcp-server": {
      "command": "uvx",
      "args": ["osm-mcp-server"]
    }
  }
}

The README specifically lists Claude Desktop, Cursor, and Windsurf as compatible hosts. For local development, install the repository in editable mode with pip install -e ., then start it with osm-mcp-server. The process listens for MCP requests on standard input and output.

Tools and capabilities

The jagan-shanmugam/open-streetmap-mcp MCP server exposes these tools:

• geocode_address and reverse_geocode for forward and reverse address lookup.
• find_nearby_places, search_category, and explore_area for place and area discovery.
• get_route_directions and analyze_commute for travel and transportation queries.
• suggest_meeting_point for selecting a meeting location for multiple people.
• find_schools_nearby for locating educational institutions.
• find_ev_charging_stations for filtered EV charging searches.
• analyze_neighborhood for neighborhood livability analysis.
• find_parking_facilities for parking searches that can include availability and fee information.

These capabilities fit applications that need geographic context without building separate MCP wrappers for each location operation.