Content
82%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.
A dense, well-organized operational skill body: exact tool names and parameter quirks, numeric quality thresholds, fallback loops, and honest limitations with near-zero padding. The main improvements are one complete example tool call and explicit per-phase validation checkpoints.
Suggestions
Add one fully-specified example tool invocation per critical tool (e.g., a complete MGnify_search_studies or GTDB_search_taxon call with all arguments) so the call shape is copy-paste ready.
Add explicit validation checkpoints between phases (e.g., 'Phase 2: if GTDB returns no hits, apply the Edge Case fallbacks before proceeding') to close the workflow_clarity gap.
Consider moving the KEGG pathway table and per-phase detail into a references/ file referenced from SKILL.md to improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and efficient throughout: compact tables for databases, quality tiers, and KEGG pathways; no explanation of concepts Claude already knows (metagenomics, CheckM, MeSH are used, not defined); every line carries operational guidance. | 5 / 5 |
Actionability | Highly concrete guidance — exact tool signatures ('GTDB_search_taxon(operation="search_taxon", query=name)'), parameter gotchas ('kegg_search_pathway(keyword=...) (NOT query)'), query syntax ('study_title="*IBD*"'), MeSH term examples, and numeric MIMAG thresholds. The gap is the absence of one complete end-to-end example tool invocation showing the full argument structure. | 4 / 5 |
Workflow Clarity | An 8-phase workflow is clearly sequenced with per-phase notes and explicit error-recovery fallbacks ('If ENA fails, fall back to MGnify', 'GMrepo 0 results: Use formal MeSH terms or NCBI taxon IDs'), plus an evidence-grading rubric for outputs. Falls short of 5 only because there are no explicit per-phase validation checkpoints (e.g., verify non-empty results before proceeding). | 4 / 5 |
Progressive Disclosure | No bundle files exist, and the single-file body is well organized with clear sections (Core Databases, Workflow, Key Phase Notes, Edge Cases, Limitations) making navigation easy. Detail-heavy material such as the KEGG pathway table and per-phase notes could live in a reference file, so the split is good but not maximal. | 4 / 5 |
Total | 17 / 20 Passed |