Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Shared support package for the Design Judge skill collection. Install it with design-award-search and design-award-match so those skills can read the canonical functional-design taxonomy and official award-source registry. Do not invoke it as a standalone design workflow.
.claude/skills/seanj1ang-design-judge-shared/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -74% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -50% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -79% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -51% | 0% |
Provide shared reference material to other Design Judge skills.
category-taxonomy.md when a workflow needs a canonical functional-design category.source-registry.md when a workflow needs official award domains, source precedence, query patterns, or verification rules.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,079 | 8,919 | +47% | 1 | 1 | 0% | 993 | 1,608 | +62% | 0 | 0 | — |
case-08 | fail→pass | 18,596 | 1,701 | -91% | 1 | 1 | 0% | 1,317 | 348 | -74% | 0 | 0 | — |
case-02 | fail→fail | 16,771 | 10,708 | -36% | 1 | 1 | 0% | 2,776 | 2,002 | -28% | 0 | 0 | — |
case-03 | fail→pass | 40,831 | 11,139 | -73% | 1 | 1 | 0% | 4,154 | 2,091 | -50% | 0 | 0 | — |
case-04 | pass→pass | 3,857 | 4,907 | +27% | 1 | 1 | 0% | 764 | 901 | +18% | 0 | 0 | — |
case-05 | pass→pass | 7,523 | 6,763 | -10% | 1 | 1 | 0% | 1,276 | 1,290 | +1% | 0 | 0 | — |
case-06 | pass→pass | 7,163 | 5,351 | -25% | 1 | 1 | 0% | 1,216 | 891 | -27% | 0 | 0 | — |
case-07 | fail→pass | 10,276 | 1,683 | -84% | 1 | 1 | 0% | 1,737 | 372 | -79% | 0 | 0 | — |
case-09 | fail→fail | 6,284 | 1,316 | -79% | 1 | 1 | 0% | 1,023 | 314 | -69% | 0 | 0 | — |
case-10 | fail→pass | 6,464 | 1,717 | -73% | 1 | 1 | 0% | 1,051 | 339 | -68% | 0 | 0 | — |
case-11 | fail→pass | 18,378 | 8,275 | -55% | 1 | 1 | 0% | 2,964 | 1,457 | -51% | 0 | 0 | — |
case-12 | fail→fail | 20,861 | 34,968 | +68% | 1 | 1 | 0% | 3,995 | 6,243 | +56% | 0 | 0 | — |
case-13 | fail→pass | 8,961 | 1,639 | -82% | 1 | 1 | 0% | 1,412 | 372 | -74% | 0 | 0 | — |
case-14 | fail→pass | 5,002 | 1,270 | -75% | 1 | 1 | 0% | 761 | 267 | -65% | 0 | 0 | — |
case-15 | fail→fail | 14,066 | 13,595 | -3% | 1 | 1 | 0% | 2,292 | 2,234 | -3% | 0 | 0 | — |
case-16 | fail→pass | 14,959 | 1,824 | -88% | 1 | 1 | 0% | 2,325 | 367 | -84% | 0 | 0 | — |
case-17 | fail→pass | 15,830 | 1,837 | -88% | 1 | 1 | 0% | 2,533 | 383 | -85% | 0 | 0 | — |
case-18 | fail→pass | 19,069 | 1,910 | -90% | 1 | 1 | 0% | 1,247 | 355 | -72% | 0 | 0 | — |
case-19 | fail→fail | 21,607 | 34,733 | +61% | 1 | 1 | 0% | 3,470 | 6,241 | +80% | 0 | 0 | — |
case-20 | fail→pass | 19,249 | 1,879 | -90% | 1 | 1 | 0% | 2,844 | 394 | -86% | 0 | 0 | — |
case-21 | fail→pass | 8,496 | 1,712 | -80% | 1 | 1 | 0% | 1,253 | 329 | -74% | 0 | 0 | — |
case-22 | fail→pass | 23,860 | 2,701 | -89% | 1 | 1 | 0% | 4,416 | 401 | -91% | 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, and 21 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 +59 percentage points is the difference between those two pass rates over the 21 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.