Content
75%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, highly actionable standards skill with excellent specific numeric guidance, a clear sequenced workflow with validation checkpoints, and a clean one-level-deep reference structure. Its main inefficiencies are the redundant Failure patterns section duplicating per-standard common failures, some re-explanation of design fundamentals, and one cross-skill reference path that sits outside the bundle.
Suggestions
Merge the 'Failure patterns' section into each standard's existing 'Common failures' list — they currently duplicate nearly the same items — and trim the 'Visual hierarchy' primer to just the standards and failure modes, cutting the design-fundamentals explanations Claude already knows.
Add an explicit error-recovery loop to the workflow (e.g., 'If a contrast check or breakpoint test fails, fix the token or layout issue and re-run the check before proceeding') to close the validation gap.
Resolve the cross-skill reference 'brand-identity/references/contrast-and-accessibility.md' — either copy the essential contrast math/checker guidance into this skill's references/ or clearly label it as a sibling-skill dependency, since it is not part of this bundle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The bulk of the body is efficient and specific (exact ratios, pixel minimums, token lists), but there is noticeable padding: the standalone "Failure patterns" section largely restates the per-standard "Common failures" lists already given, the "Visual hierarchy" section re-explains design fundamentals Claude already knows ("Size. The largest element gets noticed first. Bold beats regular"), and rhetorical flourishes ("Contrast is not a preference. It is a baseline.", "Hold the line on these and most design problems disappear") add tokens without guidance value. This sits at anchor 3 (mostly efficient, some unnecessary explanation that could be tightened) rather than anchor 4, whose example has essentially no fat, but well above anchor 2's padded-library-explanation style. | 3 / 5 |
Actionability | For an instruction-only skill, the guidance is fully actionable and specific: a WCAG ratio table, exact spacing minimums ("minimum 64px desktop, 48px mobile", "44 by 44 pixels", "at least 16px text size to prevent iOS auto-zoom"), concrete test viewports ("375 to 390 pixel viewport", "375, 768, 1024, 1440 minimum"), a 7-step workflow, and a defined output format. Per the rubric's code_vs_instruction note, absence of code is not penalized when guidance is this concrete — matching anchor 5's fully executable, common-cases-covered standard. | 5 / 5 |
Workflow Clarity | The 7-step workflow is clearly sequenced with validation checkpoints present ("Run contrast checks", "Test at viewport breakpoints", "Run the pre-ship checklist") and a conditional recovery step at the start ("If tokens are not yet defined, define a working set first"). However, there is no explicit fix-and-revalidate loop when a check fails (steps say "Confirm nothing breaks" without what to do when something does), which keeps it at anchor 4 rather than anchor 5's explicit feedback-loop example. | 4 / 5 |
Progressive Disclosure | The three referenced files (design-tokens-template.md, preship-checklist.md, tailwind-patterns.md) all exist in the bundle, are one level deep, and are clearly signaled both at point of need and in a dedicated "Reference files" section with one-line descriptions. Stack-specific patterns are correctly split out of the stack-agnostic body. The gap keeping it below anchor 5: the contrast math reference points to a sibling skill's path ("brand-identity/references/contrast-and-accessibility.md") which is not part of this skill's bundle, and the ~200-line standards core could arguably shed more to the reference files. | 4 / 5 |
Total | 16 / 20 Passed |