Geo-Perception

What it does

Geo-Perception provides a structured cognitive framework for AI agents to handle geospatial data. It operates across three levels: L1 Perception (coordinates and distance), L2 Comprehension (spatial weights and patterns), and L3 Reasoning (clustering and causal analysis).

Tools

• geo_transform_crs: Coordinate system transformation.
• geo_calculate_distance: Distance and azimuth calculations.
• geo_buffer: Generates spatial buffers.
• geo_bbox: Calculates bounding boxes.
• geo_spatial_relation: Determines topological relationships.
• geo_all_relations: Full DE-9IM topological relations.
• geo_scale_analysis: Scale inference and MAUP risk assessment.
• geo_geocode: Address to coordinate conversion (via Amap API).
• geo_reverse_geocode: Coordinate to address conversion (via Amap API).
• geo_bbox_from_places: Joint bounding box for multiple locations.
• geo_spatial_weights: Generates spatial weight matrices (Queen/Rook/KNN).
• geo_moran_global: Global Moran's I analysis.
• geo_moran_local: Local LISA (HH/LL/HL/LH) analysis.
• geo_getis_ord: Getis-Ord Gi* hotspot analysis.
• geo_semantic_label: Maps topology to semantic labels.
• geo_cluster_detect: Point (DBSCAN) and polygon spatial clustering.
• geo_od_flow: OD flow aggregation and dominant flow analysis.
• geo_causal_check: Causal premise verification using spatial autocorrelation.

Installation

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "geo-perception": {
      "command": "python",
      "args": ["path/to/geo-perception/run.py"],
      "env": {
        "PROJ_LIB": "path/to/proj/data",
        "AMAP_API_KEY": "your-amap-key"
      }
    }
  }
}

Supported hosts

• Claude