Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create distinctive, production-grade frontend interfaces with high design quality. Use when the user asks to build web components, pages, or applications and the visual direction matters as much as the code quality.
.claude/skills/ethanyoq-frontend-design/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 56% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 43% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 35% | 0% |
| case-06 | ✓→✓ | = Same ✓ | 41% | 0% |
| case-07 | ✓→✓ | = Same ✓ | 196% | 0% |
Use this when the task is not just "make it work" but "make it look designed."
This skill is for product pages, dashboards, app shells, components, or visual systems that need a clear point of view instead of generic AI-looking UI.
Pick a direction and commit to it.
Safe-average UI is usually worse than a strong, coherent aesthetic with a few bold choices.
Before coding, settle:
Possible directions:
Do not mix directions casually. Choose one and execute it cleanly.
Define:
Use CSS variables or the project's token system so the interface stays coherent as it grows.
Prefer:
Avoid defaulting to a symmetrical card grid unless it is clearly the right fit.
Use animation to:
Do not scatter generic micro-interactions everywhere. One well-directed load sequence is usually stronger than twenty random hover effects.
Use atmosphere:
Flat empty backgrounds are rarely the best answer for a product-facing page.
Never default to:
Before delivering:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,658 | 22,865 | +11% | 1 | 1 | 0% | 2,798 | 4,354 | +56% | 0 | 0 | — |
case-02 | fail→fail | 20,389 | 16,840 | -17% | 1 | 1 | 0% | 2,816 | 3,481 | +24% | 0 | 0 | — |
case-03 | fail→fail | 16,508 | 19,250 | +17% | 1 | 1 | 0% | 2,597 | 3,746 | +44% | 0 | 0 | — |
case-04 | pass→pass | 16,775 | 16,900 | +1% | 1 | 1 | 0% | 2,683 | 3,833 | +43% | 0 | 0 | — |
case-05 | pass→pass | 17,676 | 19,711 | +12% | 1 | 1 | 0% | 3,367 | 4,562 | +35% | 0 | 0 | — |
case-06 | pass→pass | 17,021 | 16,729 | -2% | 1 | 1 | 0% | 2,546 | 3,581 | +41% | 0 | 0 | — |
case-07 | pass→pass | 14,452 | 28,570 | +98% | 1 | 1 | 0% | 2,336 | 6,926 | +196% | 0 | 0 | — |
case-08 | pass→pass | 12,387 | 22,984 | +86% | 1 | 1 | 0% | 1,890 | 5,242 | +177% | 0 | 0 | — |
case-09 | pass→pass | 10,418 | 6,892 | -34% | 1 | 1 | 0% | 1,748 | 1,972 | +13% | 0 | 0 | — |
case-10 | pass→pass | 14,583 | 12,582 | -14% | 1 | 1 | 0% | 2,118 | 3,000 | +42% | 0 | 0 | — |
case-11 | pass→pass | 13,027 | 10,652 | -18% | 1 | 1 | 0% | 1,973 | 2,726 | +38% | 0 | 0 | — |
case-12 | pass→pass | 21,162 | 30,726 | +45% | 1 | 1 | 0% | 3,451 | 6,630 | +92% | 0 | 0 | — |
case-13 | pass→pass | 13,925 | 14,630 | +5% | 1 | 1 | 0% | 2,026 | 2,770 | +37% | 0 | 0 | — |
case-14 | pass→pass | 12,818 | 20,561 | +60% | 1 | 1 | 0% | 1,937 | 4,432 | +129% | 0 | 0 | — |
case-15 | pass→pass | 15,846 | 15,840 | -0% | 1 | 1 | 0% | 2,750 | 3,548 | +29% | 0 | 0 | — |
case-16 | pass→pass | 13,232 | 12,810 | -3% | 1 | 1 | 0% | 2,008 | 2,436 | +21% | 0 | 0 | — |
case-17 | pass→pass | 14,619 | 30,045 | +106% | 1 | 1 | 0% | 2,373 | 6,933 | +192% | 0 | 0 | — |
case-18 | pass→pass | 9,832 | 14,024 | +43% | 1 | 1 | 0% | 1,693 | 3,146 | +86% | 0 | 0 | — |
case-19 | pass→pass | 17,331 | 14,675 | -15% | 1 | 1 | 0% | 2,490 | 3,484 | +40% | 0 | 0 | — |
case-20 | pass→pass | 2,682 | 2,800 | +4% | 1 | 1 | 0% | 506 | 1,254 | +148% | 0 | 0 | — |
case-21 | pass→pass | 3,695 | 4,477 | +21% | 1 | 1 | 0% | 578 | 1,542 | +167% | 0 | 0 | — |
case-22 | pass→pass | 8,599 | 8,145 | -5% | 1 | 1 | 0% | 1,477 | 2,483 | +68% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.