Content
71%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 strong, opinionated instruction-only skill: concrete rules, escalation triggers, a remedial hierarchy, and an explicit output contract with decision criteria. Its two real weaknesses are the missing 'use when' posture in the frontmatter description and the three references to a STANDARDS.md file that is absent from the bundle, which undermines both the progressive-disclosure structure and the promised source of exact values.
Suggestions
Add the STANDARDS.md file (or another bundle location) so the three body references resolve; without it, exact curves/durations/spring configs the skill tells Claude to cite do not exist.
Trim rhetorical posture lines ("brutal eye for craft", "Approval is earned, not assumed") that repeat the same idea as the description.
Make the review sequence explicit (apply the ten standards → check escalation triggers → rank fixes by the remedial hierarchy → emit the two-part output), so the workflow order is stated rather than inferred from section order.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and every section carries concrete rules; there is no explanation of concepts Claude already knows. Minor rhetorical padding ("brutal eye for craft", "Approval is earned, not assumed", "Nothing appears from nothing") keeps it below the lean anchor 5. | 4 / 5 |
Actionability | Highly actionable for an instruction-only skill: exact property names (`transform`, `opacity`, `width`/`height`), concrete thresholds (300ms, `scale(0.9–0.97)`, 30–80ms stagger), a literal findings-table template with copy-ready before/after rows, and explicit Block/Approve criteria. Below 5 because the exact values (curves, duration tables, spring configs) are delegated to STANDARDS.md, which is not present in the bundle. | 4 / 5 |
Workflow Clarity | The flow is legible: ten standards to measure every animation against, escalation triggers to flag on sight, a remedial hierarchy for fixes, and a required two-part output with an explicit decision gate (Block/Approve) that functions as a validation checklist. Not 5 because the workflow is implied by section order rather than explicitly sequenced, and there is no stated step for handling ambiguous findings beyond deferring to STANDARDS.md. | 4 / 5 |
Progressive Disclosure | The body is well-sectioned and the STANDARDS.md reference is clearly signaled and one level deep ("Load it whenever a finding needs a precise value or citation") — but STANDARDS.md does not exist in the skill directory, and no references/, scripts/, or assets/ bundle is present. The disclosed second level is a dangling path, so the split is nominal rather than real. Not 2 because the inline structure itself is good and the reference is prominently signaled, not buried. | 3 / 5 |
Total | 15 / 20 Passed |