Openstreetmap MCP Server

🚀 OpenStreetMap (OSM) MCP Server

An OpenStreetMap MCP implementation that enhances the capabilities of Large Language Models (LLMs), integrating geolocation services and spatial data.

🚀 Quick Start

This server provides LLMs with tools to interact with OpenStreetMap data, enabling the following functionalities for geolocation applications:

• Convert addresses and place names to coordinates.
• Reverse geocode coordinates to addresses.
• Find nearby points of interest.
• Get route directions between places.
• Search for places within a given range.
• Display nearby locations by category.

✨ Features

This server equips LLMs with tools to interact with OpenStreetMap data, enabling the following functionalities for geolocation applications:

• Convert addresses and place names to coordinates.
• Reverse geocode coordinates to addresses.
• Find nearby points of interest.
• Get route directions between places.
• Search for places within a given range.
• Display nearby locations by category.

📺 Demo

Meeting Point Optimization

Neighborhood Analysis

Parking Search

📦 Components

Resources

Geolocation (/location)

• Geocoding: Convert natural language descriptions to latitude and longitude coordinates.
  • Request: POST /location
  • Parameters:
    • query: Natural language query, e.g., "Empire State Building".
    • bounds: Bounding box for the search area.
  • Response:

    {
      "results": [
        {
          "display_name": "Place Name",
          "lat": "Latitude",
          "lon": "Longitude",
          "confidence": "Confidence Level"
        }
      ]
    }
    

Search (/search)

• Nearby Search: Find places within a specified location and radius.
  • Request: GET /search
  • Parameters:
    • latitude: Latitude of the center point.
    • longitude: Longitude of the center point.
    • radius: Search radius in meters.
    • categories: Specified categories, e.g., ["restaurant", "hotel"].
  • Response:

    {
      "results": [
        {
          "display_name": "Place Name",
          "lat": "Latitude",
          "lon": "Longitude",
          "category": "Category"
        }
      ],
      "total_count": "Total Number of Results"
    }
    

💻 Usage Examples

Real Estate Decision-Making

• Use Case: Find suitable areas for purchasing a property in a city center.
• Steps:
  1. Use geocoding to obtain the latitude and longitude of the target area.
  2. Search for educational resources, medical facilities, and commercial centers in the area.
  3. Analyze the results to evaluate the living quality of the area.

EV Charging Station Search

• Use Case: Find the nearest electric vehicle charging station.
• Steps:
  1. Obtain the latitude and longitude of the current location.
  2. Search for places categorized as "electric_charge" within a 500 - meter radius.
  3. Display and sort the results.

⚙️ Configuration

Developer Configuration

When running the server in a development environment, add the following to the pyproject.toml file in the project root directory:

[build-system]
requires = ["uvu"]

🐞 Debugging

Since the MCP server runs via standard input and output, debugging can be challenging. We highly recommend using the MCP Inspector for debugging.

You can start the MCP Inspector with the following command:

npx @modelcontextprotocol/inspector uv --directory /path/to/osm-mcp-server run osm-mcp-server

After starting, the inspector will display a URL that you can access in your browser to begin debugging.

💻 Usage Examples

Basic Usage

import asyncio
from mcp.client import Client

async def main():
    client = Client("http://localhost:8000")
    
    # Geocoding
    results = await client.invoke_tool("geocode_address", {"address": "帝国大厦"})
    print(f"Found location: {results[0]['display_name']}")
    
    # Get coordinates
    lat = float(results[0]['lat'])
    lon = float(results[0]['lon'])
    
    # Search for nearby cafes
    nearby = await client.invoke_tool(
        "find_nearby_places",
        {
            "latitude": lat,
            "longitude": lon,
            "radius": 500,
            "categories": ["cafe"],
            "limit": 5
        }
    )
    
    # Display results
    print(f"Found {nearby['total_count']} nearby places")
    for category, subcategories in nearby["categories"].items():
        for subcategory, places in subcategories.items():
            print(f"  - {subcategory}: {len(places)} places")

if __name__ == "__main__":
    asyncio.run(main())