LocuSync Server
A geospatial MCP server that provides tools for geocoding, routing, elevation profiles, and spatial analysis. It enables AI agents to process GIS file formats like GeoJSON and Shapefiles while performing complex coordinate transformations and distance calculations.
README
LocuSync Server
A Model Context Protocol (MCP) server providing geospatial tools for AI agents. Enables Claude, GPT, and other LLMs to perform geocoding, routing, spatial analysis, and file operations.
Features
- Geocoding: Convert addresses to coordinates and vice versa (via Nominatim/OSM or Pelias)
- Batch Geocoding: Geocode multiple addresses in a single request (up to 10)
- Elevation Data: Get altitude for points and elevation profiles along paths
- Routing: Calculate routes between points with distance, duration, and geometry (via OSRM)
- Spatial Analysis: Buffer, intersection, union, distance calculations
- File I/O: Read/write Shapefiles, GeoJSON, GeoPackage
- CRS Transformation: Convert between coordinate reference systems
Installation
# From PyPI (when published)
pip install locusync-server
# From source
git clone https://github.com/matbel91765/locusync-server.git
cd locusync-server
pip install -e .
Quick Start
With Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"locusync": {
"command": "uvx",
"args": ["locusync-server"]
}
}
}
Direct Usage
# Run the server
locusync-server
Available Tools
Geocoding
geocode
Convert an address to coordinates.
Input: "1600 Pennsylvania Avenue, Washington DC"
Output: {lat: 38.8977, lon: -77.0365, display_name: "White House..."}
reverse_geocode
Convert coordinates to an address.
Input: lat=48.8566, lon=2.3522
Output: {display_name: "Paris, Île-de-France, France", ...}
batch_geocode
Geocode multiple addresses at once (max 10).
Input: addresses=["Paris, France", "London, UK", "Berlin, Germany"]
Output: {results: [...], summary: {total: 3, successful: 3, failed: 0}}
Elevation
get_elevation
Get altitude for a point.
Input: lat=48.8566, lon=2.3522
Output: {elevation_m: 35, location: {lat: 48.8566, lon: 2.3522}}
get_elevation_profile
Get elevations along a path.
Input: coordinates=[[2.3522, 48.8566], [2.2945, 48.8584]]
Output: {profile: [...], stats: {min: 28, max: 42, gain: 14}}
Geometry
distance
Calculate distance between two points.
Input: lat1=48.8566, lon1=2.3522, lat2=51.5074, lon2=-0.1278
Output: {distance: {meters: 343556, kilometers: 343.56, miles: 213.47}}
buffer
Create a buffer zone around a geometry.
Input: geometry={type: "Point", coordinates: [2.3522, 48.8566]}, distance_meters=1000
Output: {geometry: {type: "Polygon", ...}, area_km2: 3.14}
spatial_query
Perform spatial operations (intersection, union, contains, within, etc.).
Input: geometry1={...}, geometry2={...}, operation="intersection"
Output: {geometry: {...}}
transform_crs
Transform coordinates between CRS.
Input: geometry={...}, source_crs="EPSG:4326", target_crs="EPSG:3857"
Output: {geometry: {...}}
Routing
route
Calculate route between two points.
Input: start_lat=48.8566, start_lon=2.3522, end_lat=48.8606, end_lon=2.3376
Output: {distance: {...}, duration: {...}, geometry: {...}, steps: [...]}
isochrone
Calculate area reachable within a time limit.
Input: lat=48.8566, lon=2.3522, time_minutes=15, profile="driving"
Output: {geometry: {type: "Polygon", ...}}
Files
read_file
Read geospatial files (Shapefile, GeoJSON, GeoPackage).
Input: file_path="data/cities.shp"
Output: {type: "FeatureCollection", features: [...]}
write_file
Write features to geospatial files.
Input: features={...}, file_path="output.geojson", driver="GeoJSON"
Output: {file_path: "...", feature_count: 10}
Configuration
Environment variables:
| Variable | Default | Description |
|---|---|---|
NOMINATIM_URL |
https://nominatim.openstreetmap.org |
Nominatim API URL |
NOMINATIM_USER_AGENT |
locusync-server/1.0.0 |
User agent for Nominatim |
OSRM_URL |
https://router.project-osrm.org |
OSRM API URL |
OSRM_PROFILE |
driving |
Default routing profile |
PELIAS_URL |
(empty) | Pelias geocoding API URL |
PELIAS_API_KEY |
(empty) | Pelias API key (optional) |
OPEN_ELEVATION_URL |
https://api.open-elevation.com |
Open-Elevation API URL |
GIS_DEFAULT_CRS |
EPSG:4326 |
Default CRS |
GIS_TEMP_DIR |
/tmp/locusync |
Temporary directory |
Response Format
All tools return a consistent JSON structure:
{
"success": true,
"data": { ... },
"metadata": {
"source": "nominatim",
"confidence": 0.95
},
"error": null
}
Rate Limits
- Nominatim: 1 request/second (enforced automatically)
- OSRM Demo: Best effort, consider self-hosting for production
Development
# Install dev dependencies
pip install -e ".[dev]"
# Run tests
pytest
# Run with coverage
pytest --cov=src/locusync --cov-report=html
# Type checking
mypy src/locusync
# Linting
ruff check src/locusync
Architecture
src/locusync/
├── server.py # MCP server entry point
├── config.py # Configuration management
├── utils.py # Common utilities
└── tools/
├── geocoding.py # geocode, reverse_geocode, batch_geocode
├── elevation.py # get_elevation, get_elevation_profile
├── routing.py # route, isochrone
├── geometry.py # buffer, distance, spatial_query, transform_crs
└── files.py # read_file, write_file
License
MIT License - see LICENSE for details.
Contributing
Contributions welcome! Please read the contributing guidelines before submitting PRs.
Roadmap
- [x] Pelias geocoding support (higher accuracy)
- [x] Elevation/terrain data
- [x] Batch geocoding
- [ ] Valhalla routing integration (native isochrones)
- [ ] PostGIS spatial queries
- [ ] Real-time traffic data
- [ ] ESRI FileGDB full support
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。