Content
73%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 well-structured instruction skill with an excellent reference-module architecture, explicit workflow checkpoints, and concrete output specifications. Its main weaknesses are redundancy across the Core Function / Hard Rules / What This Skill Should Not Do sections (partially duplicating both each other and the reference modules) and the absence of a worked output example.
Suggestions
Merge 'Sample Triggers' into 'Input Validation' and consolidate the overlapping prohibitions from 'Core Function', 'Hard Rules', and 'What This Skill Should Not Do' into a single anti-patterns section, trusting the reference modules to carry the detailed rules.
Remove the inline design-family list in Step 2 (it duplicates references/study-design-routing-rules.md) and the rules in 'Hard Rules' that restate references/literature-integrity-rules.md, keeping only pointers.
Add a short worked example (e.g., one filled-in Section F workflow step and one Section I gap entry from a real paper) so the output format is unambiguous without reading output-section-guidance.md.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The text is dense with actionable rules and free of filler explaining known concepts, but redundancy spans multiple sections: trigger examples appear in both 'Input Validation' and 'Sample Triggers', prohibitions on inventing details appear in 'Core Function', 'Hard Rules' (rules 7, 13-14), and 'What This Skill Should Not Do', and the explicit/inferred distinction is stated twice. This is 'mostly efficient but could be tightened' rather than padded with unnecessary explanation. | 3 / 5 |
Actionability | Guidance is highly concrete and executable for an instruction-only skill: ten sequenced execution steps, eleven output sections with per-section content requirements, a copy-paste redirect script for out-of-scope requests, and explicit labeling vocabulary ('explicitly reported / strongly inferable / missing'). A minor gap keeps it below 5: no worked example of the output format (e.g., a sample Section F workflow entry). | 4 / 5 |
Workflow Clarity | Steps 1-10 are clearly sequenced and mapped to output sections A-K, with explicit checkpoints: coverage gating ('Use the coverage rules... before attempting full reconstruction'), a stop-and-redirect rule for out-of-scope inputs, a dedicated audit step (Step 9), and a completeness validation rule ('If any output section is generated without using its corresponding reference module, the output should be treated as incomplete'). | 5 / 5 |
Progressive Disclosure | All 11 reference files exist, are one level deep, and are clearly signaled with per-module use points ('references/study-design-routing-rules.md → use when identifying the dominant design family before reconstruction in Section B'), repeated at the relevant execution steps. Minor organization gap: the body re-inlines content that lives in the reference modules — Step 2's design-family list duplicates study-design-routing-rules.md and the 17 Hard Rules overlap literature-integrity-rules.md. | 4 / 5 |
Total | 16 / 20 Passed |