Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Enforce Material Design 3 theming and design token usage in Jetpack Compose. Use when implementing M3 components, color schemes, typography, or design tokens.
.claude/skills/hoangnguyen0403-android-design-system/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 113% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 50% | 0% |
Define Color.kt, Theme.kt, and Type.kt in ui/theme/. Map every raw color/type value to lightColorScheme/darkColorScheme slots. Access all tokens through MaterialTheme:
MaterialTheme.colorScheme.*MaterialTheme.typography.*.dp units consistentlyMaterialTheme.colorScheme.*, not Color(0xFF...).MaterialTheme.typography.*, not raw fontSize = 32.sp..dp tokens; avoid unexplained magic numbers.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 18,411 | 16,093 | -13% | 1 | 1 | 0% | 4,129 | 4,118 | -0% | 0 | 0 | — |
case-02 | fail→pass | 17,989 | 21,638 | +20% | 1 | 1 | 0% | 3,751 | 5,366 | +43% | 0 | 0 | — |
case-03 | fail→pass | 10,268 | 8,736 | -15% | 1 | 1 | 0% | 1,833 | 2,016 | +10% | 0 | 0 | — |
case-22 | pass→pass | 16,439 | 17,596 | +7% | 1 | 1 | 0% | 3,561 | 4,422 | +24% | 0 | 0 | — |
case-04 | fail→pass | 8,530 | 17,317 | +103% | 1 | 1 | 0% | 2,042 | 4,359 | +113% | 0 | 0 | — |
case-05 | fail→pass | 15,540 | 22,217 | +43% | 1 | 1 | 0% | 3,459 | 5,182 | +50% | 0 | 0 | — |
case-06 | pass→pass | 13,750 | 23,884 | +74% | 1 | 1 | 0% | 3,117 | 4,876 | +56% | 0 | 0 | — |
case-07 | pass→pass | 12,448 | 9,859 | -21% | 1 | 1 | 0% | 2,745 | 2,500 | -9% | 0 | 0 | — |
case-08 | pass→pass | 12,834 | 11,727 | -9% | 1 | 1 | 0% | 2,546 | 2,537 | -0% | 0 | 0 | — |
case-09 | fail→pass | 13,972 | 17,611 | +26% | 1 | 1 | 0% | 2,963 | 4,274 | +44% | 0 | 0 | — |
case-10 | fail→pass | 8,861 | 10,619 | +20% | 1 | 1 | 0% | 1,998 | 2,654 | +33% | 0 | 0 | — |
case-11 | fail→pass | 9,106 | 8,863 | -3% | 1 | 1 | 0% | 1,983 | 2,153 | +9% | 0 | 0 | — |
case-12 | fail→fail | 10,964 | 19,149 | +75% | 1 | 1 | 0% | 2,435 | 3,636 | +49% | 0 | 0 | — |
case-13 | fail→pass | 6,117 | 10,366 | +69% | 1 | 1 | 0% | 1,381 | 2,470 | +79% | 0 | 0 | — |
case-14 | fail→pass | 7,548 | 14,354 | +90% | 1 | 1 | 0% | 1,740 | 3,525 | +103% | 0 | 0 | — |
case-15 | fail→pass | 14,692 | 16,183 | +10% | 1 | 1 | 0% | 3,012 | 4,117 | +37% | 0 | 0 | — |
case-16 | pass→pass | 14,980 | 13,264 | -11% | 1 | 1 | 0% | 3,114 | 3,173 | +2% | 0 | 0 | — |
case-17 | fail→pass | 9,231 | 17,448 | +89% | 1 | 1 | 0% | 2,019 | 4,347 | +115% | 0 | 0 | — |
case-18 | fail→pass | 6,243 | 13,180 | +111% | 1 | 1 | 0% | 1,373 | 3,108 | +126% | 0 | 0 | — |
case-19 | fail→fail | 11,292 | 10,017 | -11% | 1 | 1 | 0% | 2,005 | 2,010 | +0% | 0 | 0 | — |
case-20 | pass→pass | 10,250 | 9,420 | -8% | 1 | 1 | 0% | 2,137 | 2,169 | +1% | 0 | 0 | — |
case-21 | pass→pass | 10,179 | 11,500 | +13% | 1 | 1 | 0% | 2,058 | 2,649 | +29% | 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 +59 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.