Open Streetmap MCP

🚀 OpenStreetMap (OSM) MCP Server

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

🚀 Quick Start

This server combines location services with geospatial data to enhance the capabilities of LLMs. It offers various features such as geocoding, reverse geocoding, nearby place search, route planning, and access to OpenStreetMap's raw geographic data.

🔍 Demo

Meeting Point Optimization

Community Analysis

Parking Lot Search

✨ Features

• Geocoding: Convert addresses to latitude and longitude.
• Reverse Geocoding: Obtain address information based on latitude and longitude.
• Nearby Place Search: Find points of interest within a specified range.
• Route Planning: Calculate the optimal route between two points.
• Map Data Access: Access the raw geographic data of OpenStreetMap.

🧩 Components

Resources

1. Location Resources: Provide detailed information about locations, including latitude, longitude, address, and relevant tags.
2. Search Resources: Support location searches based on keywords, categories, or filtering conditions.
3. Route Planning Resources: Calculate the path between two points and return detailed path data.

Tools

• Geocoding Tool: Used to convert addresses to latitude and longitude.
• Reverse Geocoding Tool: Obtain address information based on latitude and longitude.
• Nearby Place Tool: Find points of interest within a specified range.
• Route Planning Tool: Calculate the optimal route between two points.

📋 Use Cases

Real Estate Decision-Making

Assist users in analyzing the real estate value of a target area and evaluating property prices based on surrounding facilities and transportation convenience.

EV Charging Station Search

Provide electric vehicle users with information on the location, availability, and real-time status of nearby charging stations.

💻 Usage Examples

Local Testing

Running the Server

1. Install dependencies:

uv sync

2. Start the server:

uv run --port 8000

Testing the Client

import asyncio
from mcp.client import Client

async def main():
    client = Client("http://localhost:8000")
    
    # Geocoding test
    results = await client.invoke_tool("geocode_address", {"address": "Empire State Building"})
    print(f"Location found: {results[0]['display_name']}")
    
    # Get coordinates
    lat = float(results[0]['lat'])
    lon = float(results[0]['lon'])
    
    # Find nearby cafes
    nearby = await client.invoke_tool(
        "find_nearby_places",
        {
            "latitude": lat,
            "longitude": lon,
            "radius": 500,
            "categories": ["amenity"],
            "limit": 5
        }
    )
    
    # Print 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())

Configuration

Development Mode

1. Install dependencies:

uv sync

2. Start the server:

uv run --port 8000

Production Environment

1. Build the release package:

uv build

2. Deploy to the production environment and ensure the correct path of the configuration file:

• MacOS: ~/Library/Application Support/osm-mcp-server/
• Windows: %APPDATA%/osm-mcp-server/
• Linux: ~/.config/osm-mcp-server/

Development and Release

Preparing for Release

1. Synchronize dependencies:

uv sync

2. Build the release package:

uv build

3. Publish to PyPI:

uv publish

Debugging

Use the MCP insulator tool for debugging:

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

Example API Usage

import asyncio
from mcp.client import Client

async def main():
    client = Client("http://localhost:8000")
    
    # Geocoding test
    results = await client.invoke_tool("geocode_address", {"address": "Empire State Building"})
    print(f"Location found: {results[0]['display_name']}")
    
    # Get coordinates
    lat = float(results[0]['lat'])
    lon = float(results[0]['lon'])
    
    # Find nearby cafes
    nearby = await client.invoke_tool(
        "find_nearby_places",
        {
            "latitude": lat,
            "longitude": lon,
            "radius": 500,
            "categories": ["amenity"],
            "limit": 5
        }
    )
    
    # Print 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())