Content
65%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is highly actionable with concrete checklists, sequenced procedures, and an output template, but it is hurt by redundancy across sections, missing validation checkpoints for destructive database operations, and reliance on a 'rules/' directory and 'AGENTS.md' that are not present in the bundle.
Suggestions
Ship the referenced 'rules/*.md' files and 'AGENTS.md' (or remove the references) so progressive-disclosure navigation resolves to real content.
Add an explicit validation step to the review workflow, e.g. 'Confirm each finding cites a real rule and verify the suggested change is non-destructive before recommending it', to satisfy the destructive-operation feedback-loop requirement.
Dedupe the rule listings — keep either 'Review Procedures' or 'Quick Reference' as the primary index and cross-reference rather than repeating every rule name with its description twice.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient actionable reference material, but rule names and counts recur redundantly across 'Review Procedures', the priority table, and 'Quick Reference' (e.g. schema-pk and insert-mutation rules each appear 2-3 times), so it could be tightened. It does not earn a 3 because of that repetition; it avoids a 1 by not explaining basic concepts Claude already knows. | 2 / 3 |
Actionability | Concrete sequenced procedures ('Read these rule files in order'), specific quantitative checklists ('Batch size 10K-100K rows', 'partition count 100-1,000'), and a copy-paste output format template provide fully executable guidance. | 3 / 3 |
Workflow Clarity | A clear priority order and sequenced rule reading exist, but this is a database-review skill covering destructive/batch operations (ALTER TABLE, mutations, OPTIMIZE) and lacks any validation or verification checkpoint before issuing recommendations, which per the rubric caps workflow clarity at 2. | 2 / 3 |
Progressive Disclosure | The body is well-structured with one-level-deep references to a 'rules/' directory and 'AGENTS.md', but none of these referenced bundle files actually exist in the skill, so the navigation points to missing material rather than genuinely splitting content. It is above 1 because the in-body structure is clear and well-signaled, but below 3 because the references are broken. | 2 / 3 |
Total | 9 / 12 Passed |