Content
92%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.
This is a well-engineered instruction skill: a clearly sequenced ten-step workflow with per-step validation CLIs, a release gate, and clean one-level-deep reference structure where every cited bundle file exists. Its only notable weakness is conciseness — a few inline lists (survey disclosure fields, ethics "Never" bullets, the citation appendix) duplicate material that belongs in the already-bundled references.
Suggestions
Move the section-7 survey/interview disclosure field lists into references/methods_and_ethics.md, keeping only the top-level rules inline, since that reference already covers the same ground.
Trim or relocate the closing "Citing Scientific Agent Skills" section to a short reference file, leaving a one-line citation reminder in SKILL.md.
Condense the section-1 research-contract bullet list into the few decisions that materially change the denominator, pointing to references/report_structure_guide.md for the full checklist.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense, assumes domain competence (never explains what TAM/SAM/SOM or a survey is), and almost every list encodes non-obvious procedural knowledge (ledger fields, coverage guardrails, reconciliation rules). Minor trimming opportunities remain — e.g., the section-7 survey-disclosure field list and the "Never" bullets overlap with references/methods_and_ethics.md, and the long citation appendix could live in a reference — so it fits "efficient; minor instances that could be trimmed" rather than "every token earns its place". | 4 / 5 |
Actionability | Guidance is fully executable: copy-paste commands with concrete paths for each validator/calculator (e.g., `python3 scripts/audit_claim_citations.py data/claims.csv data/source_ledger.csv`, `python3 scripts/generate_report_scaffold.py assets/report_manifest_template.json ./market-report-workspace`), exact sizing formulas, and explicit ledger schemas. The scaffold command's constraints (horizon derivation, empty-ledger caveat) are stated. | 5 / 5 |
Workflow Clarity | A ten-step numbered workflow with explicit sequencing, per-step validation commands, a release-gate checklist, and explicit feedback-loop handling ("generated empty ledgers and analysis files must be populated before their validators can pass", re-checking API terms before retrieval). This matches the anchor of clear sequence with explicit validation steps and error-recovery guidance. | 5 / 5 |
Progressive Disclosure | SKILL.md is a genuine overview: each step points to exactly one level of references (all seven referenced files in references/ exist on disk, as do all cited assets/ and scripts/ paths), and the Bundled resources section catalogs each file with a one-line description. Content is appropriately split with clear navigation and no nested-reference chains. | 5 / 5 |
Total | 19 / 20 Passed |