Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create distinctive, production-grade frontend interfaces.
.claude/skills/hezaohezao-frontend-design/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-21 | ✓→✗ | ▼ Worse | 320% | 0% |
| case-22 | ✓→✗ | ▼ Worse | 13% | 0% |
| case-23 | ✓→✗ | ▼ Worse | 323% | 0% |
| case-01 | ✗→✗ | = Same ✗ | 11% | 0% |
Create distinctive, production-grade frontend interfaces that avoid generic "AI slop" aesthetics. Implement real working code with exceptional attention to aesthetic details and creative choices.
MANDATORY: The entry HTML file MUST be named index.html.
Before coding, understand the context and commit to a BOLD aesthetic direction:
retro-futuristic, organic/natural, luxury/refined, playful/toy-like, editorial/magazine, brutalist/raw, art deco/geometric, soft/pastel, industrial/utilitarian, etc.
someone will remember?
CRITICAL: Choose a clear conceptual direction and execute it with precision. Bold maximalism and refined minimalism both work — the key is intentionality.
Then implement working code (HTML/CSS/JS, React, Vue, etc.) that is:
reveals creates more delight than scattered micro-interactions
geometric patterns, layered transparencies, dramatic shadows, decorative borders, custom cursors, grain overlays
NEVER use generic AI-generated aesthetics:
Match implementation complexity to the aesthetic vision. Maximalist designs need elaborate code. Minimalist designs need restraint, precision, and careful attention to spacing, typography, and subtle details.
write_file("index.html", "<full HTML content>") write_file("styles.css", "<full CSS content>") write_file("script.js", "<full JS content>")
AI-generated page, redo it with a bolder direction.
moment beats ten scattered micro-interactions.
contrast ratios, font sizes, focus states.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 57,810 | 42,955 | -26% | 1 | 1 | 0% | 8,265 | 9,139 | +11% | 0 | 0 | — |
case-02 | fail→fail | 89,070 | 49,423 | -45% | 1 | 1 | 0% | 8,251 | 9,741 | +18% | 0 | 0 | — |
case-03 | fail→fail | 116,588 | 139,878 | +20% | 1 | 1 | 0% | 8,698 | 3,832 | -56% | 0 | 0 | — |
case-04 | fail→fail | 83,591 | 128,349 | +54% | 1 | 1 | 0% | 8,235 | 9,900 | +20% | 0 | 0 | — |
case-05 | fail→pass | 40,593 | 100,754 | +148% | 1 | 1 | 0% | 8,230 | 8,740 | +6% | 0 | 0 | — |
case-06 | fail→fail | 36,211 | 45,157 | +25% | 1 | 1 | 0% | 8,232 | 9,106 | +11% | 0 | 0 | — |
case-07 | fail→fail | 42,533 | 44,877 | +6% | 1 | 1 | 0% | 8,500 | 9,651 | +14% | 0 | 0 | — |
case-08 | fail→fail | 61,189 | 70,590 | +15% | 1 | 1 | 0% | 8,337 | 9,100 | +9% | 0 | 0 | — |
case-09 | fail→fail | 40,344 | 92,091 | +128% | 1 | 1 | 0% | 8,215 | 11,206 | +36% | 0 | 0 | — |
case-10 | fail→fail | 42,092 | 89,882 | +114% | 1 | 1 | 0% | 8,242 | 3,178 | -61% | 0 | 0 | — |
case-11 | fail→fail | 40,122 | 42,532 | +6% | 1 | 1 | 0% | 8,227 | 9,101 | +11% | 0 | 0 | — |
case-12 | fail→fail | 71,309 | 137,482 | +93% | 1 | 1 | 0% | 8,376 | 9,102 | +9% | 0 | 0 | — |
case-13 | fail→fail | 39,114 | 43,853 | +12% | 1 | 1 | 0% | 8,229 | 9,103 | +11% | 0 | 0 | — |
case-14 | fail→fail | 49,025 | 118,190 | +141% | 1 | 1 | 0% | 8,225 | 4,097 | -50% | 0 | 0 | — |
case-15 | fail→fail | 37,924 | 65,052 | +72% | 1 | 1 | 0% | 8,239 | 8,808 | +7% | 0 | 0 | — |
case-16 | fail→fail | 36,753 | 45,617 | +24% | 1 | 1 | 0% | 8,215 | 9,089 | +11% | 0 | 0 | — |
case-17 | fail→fail | 34,586 | 61,218 | +77% | 1 | 1 | 0% | 8,479 | 12,506 | +47% | 0 | 0 | — |
case-18 | fail→fail | 69,802 | 50,798 | -27% | 1 | 1 | 0% | 9,027 | 5,338 | -41% | 0 | 0 | — |
case-19 | fail→fail | 44,639 | 47,940 | +7% | 1 | 1 | 0% | 9,592 | 3,537 | -63% | 0 | 0 | — |
case-20 | fail→fail | 59,605 | 45,212 | -24% | 1 | 1 | 0% | 8,228 | 9,102 | +11% | 0 | 0 | — |
case-21 | pass→fail | 15,622 | 67,839 | +334% | 1 | 1 | 0% | 3,132 | 13,139 | +320% | 0 | 0 | — |
case-22 | pass→fail | 13,495 | 72,221 | +435% | 1 | 1 | 0% | 2,454 | 2,769 | +13% | 0 | 0 | — |
case-23 | pass→fail | 11,324 | 41,381 | +265% | 1 | 1 | 0% | 2,153 | 9,109 | +323% | 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. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -17 percentage points is the difference between those two pass rates over the 22 comparable cases. 7 cases got worse with the skill loaded, and they are included in that figure.
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.