Content
27%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill contains genuinely valuable and opinionated design knowledge, but it suffers from significant verbosity — explaining concepts Claude already understands, repeating anti-patterns across multiple sections, and including redundant 'tips' and 'pitfalls' sections. The output template is the strongest element, providing a concrete artifact structure, but the lack of validation steps and monolithic organization weaken its effectiveness as an actionable skill.
Suggestions
Cut the content by 40-50%: remove 'Overview' explanation of what Stitch is, eliminate 'Tips for Success' and 'Common Pitfalls' sections (they restate rules already given), and merge redundant anti-pattern mentions into a single canonical list.
Add a validation checklist step after generation: e.g., 'Verify: ☐ Only 1 accent color ☐ No banned fonts ☐ All colors have hex codes ☐ Anti-patterns section populated ☐ Responsive rules included'.
Split into multiple files: keep SKILL.md as a concise overview with the output template, and move detailed rules (color constraints, typography rules, component specs, anti-patterns) into separate reference files like COLOR_RULES.md, TYPOGRAPHY.md, ANTI_PATTERNS.md.
Add a complete worked example: show a real user prompt (e.g., 'minimalist finance dashboard') and the full generated DESIGN.md output, so Claude can pattern-match rather than interpret abstract instructions.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~250+ lines. It over-explains concepts Claude already knows (what Stitch is, what a design system is, what CSS Grid vs Flexbox is), includes lengthy 'Tips for Success' and 'Common Pitfalls' sections that restate earlier content, and the 'Best Practices' section is largely redundant. The atmosphere description methodology, while novel, could be conveyed in half the tokens. | 1 / 3 |
Actionability | The skill provides a concrete output template (DESIGN.md structure) with specific hex codes, font names, and CSS values, which is good. However, it lacks executable code or commands — there's no step-by-step generation process, no script to run, and the 'Analysis & Synthesis Instructions' are descriptive rather than procedural. The output format example is helpful but is itself a template rather than a fully worked example. | 2 / 3 |
Workflow Clarity | The numbered sections (1-9) provide a clear sequence for generating the DESIGN.md, but there are no validation checkpoints — no way to verify the generated output meets the constraints, no feedback loop for checking anti-patterns are actually excluded, and no verification step before delivering the final DESIGN.md. For a generative skill producing a complex artifact, validation steps would be important. | 2 / 3 |
Progressive Disclosure | The entire skill is a monolithic wall of text with no references to supporting files. The detailed component stylings, anti-patterns list, typography rules, and motion philosophy could each be separate reference files. With no bundle files provided and everything inline, the content is poorly structured for progressive disclosure — a reader must consume everything at once. | 1 / 3 |
Total | 6 / 12 Passed |