GIS Knowledge Hub V5.0

Problems It Targets

GIS workflows often break on CRS mismatches, format rules, national standards, and toolchain differences: Shapefile, GDB, DWG, and KML each expect different coordinate systems; GB/T and industry handover rules are easy to mix with default parameters; scripts can finish without logs, checks, or acceptance evidence. This skill organizes GIS knowledge, execution constraints, atomic skills, and self-evolving feedback into a traceable workflow.

How It Works

• Layered structure: knowledge base, atomic skills, multi-agent orchestration, and GeoEvolve feedback, keeping the pipeline fixed instead of producing scattered snippets.
• Hard constraints: source data is read-only; outputs go to output_YYYYMMDD/; ambiguous units, tolerances, field names, or CRS choices trigger confirmation; automatic repair stops after three failed rounds and leaves a human todo list.
• Key steps: requirements lock, data profiling, plan generation, environment prep, core execution, quality checks, and archived delivery, with runnable code, self-check scripts, and acceptance checklists.
• Coverage: CGCS2000, Gauss-Kruger, EPSG:4326, GB/T, ArcGIS Pro, QGIS, FME, LiDAR360, GDAL, GeoPandas, PostGIS, WebGIS, and 3D GIS.

Boundaries

Best for data conversion, CRS conversion, topology/geometry QC, GB/T compliance, CAD-GIS conversion, scripted GIS processing, and delivery acceptance; not a replacement for full GIS software. Sensitive data, local standards, field surveying, or industry-specific rules still need human confirmation and compliance review.