Content
63%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 body is a well-organized, domain-rich pipeline reference with concrete tool input/output specs, edge-case fallbacks, and a useful decision matrix. Its main weaknesses are verbosity (redundant resolution-tier statements and a generic EM-modality explainer), an inconsistent phase numbering between the overview diagram and phase details, and heavy inlining of content (per-database tool references, the IDR guide) that should live in separate reference files.
Suggestions
Fix the phase-numbering inconsistency: the Workflow Overview ends at 'Phase 6: Report Synthesis' while the Phase Details contain seven phases with a different Phase 6 ('Interpretation & Recommendations') — align the diagram with the detail sections.
Cut the generic EM-modality explainer ('TEM resolves individual protein complexes (~2nm)... SEM shows surface topology') and consolidate the resolution-tier tables, which currently appear in the guiding principles, Phase 1, and the Phase 6 decision matrix.
Move separable content into one-level-deep reference files (e.g., references/IDR.md for the light-microscopy/high-content-screening section and per-database tool details), keeping SKILL.md as a concise overview with clearly signaled links.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient reference material, but padded in places: 'EM resolution determines what you can see. TEM resolves individual protein complexes (~2nm). Cryo-EM achieves near-atomic resolution (<4Å). SEM shows surface topology' re-explains EM modalities Claude already knows, and resolution tiers are stated three times (guiding principles, Phase 1 'Resolution interpretation', Phase 6 decision matrix). Matches 'mostly efficient but includes some unnecessary explanation or could be tightened'; not 2 because most content is non-obvious domain specifics (FSC 0.143 vs 0.5 cutoffs, cross-correlation > 0.7). | 3 / 5 |
Actionability | Every tool is listed with concrete inputs and outputs (e.g., 'EMDB_search_structures — Input: query, optional resolution_min, resolution_max, method, limit'), and the IDR section includes a runnable 'tu.run_tool("IDR_search_studies", {"key": "Gene Symbol", "value": "TP53"})' example. Not 5 because the EM tools themselves lack any executable call examples — only IDR has one — and per-phase workflows are prose step lists; not 3 because the input/output specs make the guidance mostly executable. | 4 / 5 |
Workflow Clarity | A clearly sequenced Phase 0–7 pipeline with per-phase objectives, tools, and workflows, plus 'Edge Cases & Fallbacks' and a Phase 6 quality-assessment checklist acting as validation checkpoints (all operations are read-only, so the destructive/batch cap does not apply). Not 5 because the Workflow Overview diagram lists only six phases ending at 'Phase 6: Report Synthesis' while the Phase Details contain seven phases with a different Phase 6 ('Interpretation & Recommendations'), an inconsistency that muddles the sequence. | 4 / 5 |
Progressive Disclosure | No bundle files exist (no references/, scripts/, or assets/), and the ~290-line body inlines content that clearly belongs in separate one-level-deep files — the per-database tool/API detail for each phase and the entire 'Related Resource: IDR' section. Not 2 because clear headers, tables, and the 'Common Analysis Patterns' map keep it well-organized and navigable; not 4 because substantial separable content sits inline in SKILL.md. | 3 / 5 |
Total | 14 / 20 Passed |