Content
53%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The content is actionable with many executable examples and a sound reference structure, but it is overlong and padded with general knowledge Claude already has, and its workflows lack the validation checkpoints expected for destructive/batch geospatial operations.
Suggestions
Cut the generic Performance Tips and Best Practices sections, or move them into a reference file; keep only geospatial-specific guidance Claude would not already know.
Add explicit validation checkpoints to the land-cover and flood workflows (e.g., verify nodata/CRS, re-open and sanity-check the written raster before proceeding).
Fix the broken references: create references/coordinate-systems.md and references/hydrology.md, or repoint the body links to existing files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | At ~680 lines the body is noticeably verbose, with padded sections explaining concepts Claude already knows (CRS fundamentals, generic Performance Tips and Best Practices lists) that inflate the token budget. | 2 / 5 |
Actionability | Mostly executable code blocks for NDVI, GeoPandas joins, GEE time series, classification, terrain, and routing, but minor gaps remain (flood workflow is a comment pseudocode block, suitability references undefined x,y and rasterize_features). | 4 / 5 |
Workflow Clarity | Multi-step workflows are present and numbered, but validation/verification checkpoints are absent for batch raster-writing and destructive operations, which caps this score per the rubric. | 3 / 5 |
Progressive Disclosure | Good structure with a Detailed Documentation section signaling 12 one-level-deep reference files, but two referenced paths (coordinate-systems.md, hydrology.md) do not exist, slightly weakening navigation. | 4 / 5 |
Total | 13 / 20 Passed |