Under heavy development by h4ck1ng.scienceReal-time collaborative canvasGeo-anchored cursors and presencePMTiles, COG, GeoParquet, GeoJSON, ZarrData read in place over HTTP range requestsDrop-to-clip: vectors filtered, rasters maskedGeodesic measurements and pinned notesShared timeline and sequencerSpatial SQL in the browser on DuckDB-WASMSPARQL layers3D modeMCP and Python agents, free on every planOffline-firstSelf-hostableNo telemetryUnder heavy development by h4ck1ng.scienceReal-time collaborative canvasGeo-anchored cursors and presencePMTiles, COG, GeoParquet, GeoJSON, ZarrData read in place over HTTP range requestsDrop-to-clip: vectors filtered, rasters maskedGeodesic measurements and pinned notesShared timeline and sequencerSpatial SQL in the browser on DuckDB-WASMSPARQL layers3D modeMCP and Python agents, free on every planOffline-firstSelf-hostableNo telemetry
Connect a Python script to Driades via MCP. The agent watches the canvas, classifies features, and draws polygons — while your team reviews each result live.