Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Design search experiences that help users find what they need, recover from failure, and refine results.
.claude/skills/owl-listener-search-ux/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-08 | ✓→✗ | ▼ Worse | 37% | 0% |
You are an expert in designing search systems that are fast, forgiving, and genuinely useful.
You design the full search experience — query input, results, filtering, zero-results states, and search-as-navigation patterns — so users find what they need with minimal effort and friction.
Some users use search to navigate rather than find: "settings", "invoices", "my profile". Design for this:
Each result should show:
The most critical and most often neglected state:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 12,901 | 15,317 | +19% | 1 | 1 | 0% | 2,087 | 3,269 | +57% | 0 | 0 | — |
case-06 | fail→pass | 19,073 | 15,649 | -18% | 1 | 1 | 0% | 2,465 | 3,234 | +31% | 0 | 0 | — |
case-01 | fail→pass | 23,531 | 27,272 | +16% | 1 | 1 | 0% | 3,653 | 5,229 | +43% | 0 | 0 | — |
case-02 | fail→fail | 16,434 | 17,366 | +6% | 1 | 1 | 0% | 2,559 | 3,506 | +37% | 0 | 0 | — |
case-03 | fail→fail | 14,045 | 15,685 | +12% | 1 | 1 | 0% | 2,047 | 3,266 | +60% | 0 | 0 | — |
case-04 | pass→pass | 15,361 | 12,013 | -22% | 1 | 1 | 0% | 2,172 | 2,553 | +18% | 0 | 0 | — |
case-07 | fail→pass | 27,010 | 14,234 | -47% | 1 | 1 | 0% | 2,171 | 3,002 | +38% | 0 | 0 | — |
case-08 | pass→fail | 15,127 | 13,719 | -9% | 1 | 1 | 0% | 2,120 | 2,899 | +37% | 0 | 0 | — |
case-09 | pass→pass | 11,198 | 12,677 | +13% | 1 | 1 | 0% | 1,816 | 2,797 | +54% | 0 | 0 | — |
case-10 | pass→pass | 13,758 | 10,403 | -24% | 1 | 1 | 0% | 2,005 | 2,397 | +20% | 0 | 0 | — |
case-11 | pass→pass | 12,145 | 15,888 | +31% | 1 | 1 | 0% | 1,824 | 3,315 | +82% | 0 | 0 | — |
case-12 | pass→pass | 15,527 | 16,630 | +7% | 1 | 1 | 0% | 2,359 | 3,387 | +44% | 0 | 0 | — |
case-13 | pass→pass | 8,072 | 10,687 | +32% | 1 | 1 | 0% | 1,350 | 2,518 | +87% | 0 | 0 | — |
case-14 | fail→pass | 10,740 | 10,577 | -2% | 1 | 1 | 0% | 1,806 | 2,625 | +45% | 0 | 0 | — |
case-15 | pass→pass | 14,456 | 13,947 | -4% | 1 | 1 | 0% | 2,440 | 3,145 | +29% | 0 | 0 | — |
case-16 | pass→pass | 13,375 | 12,272 | -8% | 1 | 1 | 0% | 2,205 | 3,008 | +36% | 0 | 0 | — |
case-21 | pass→pass | 15,573 | 17,745 | +14% | 1 | 1 | 0% | 2,953 | 3,660 | +24% | 0 | 0 | — |
case-17 | pass→pass | 5,720 | 4,820 | -16% | 1 | 1 | 0% | 909 | 1,665 | +83% | 0 | 0 | — |
case-18 | pass→pass | 5,537 | 5,607 | +1% | 1 | 1 | 0% | 906 | 1,784 | +97% | 0 | 0 | — |
case-19 | pass→pass | 14,785 | 14,522 | -2% | 1 | 1 | 0% | 2,345 | 3,162 | +35% | 0 | 0 | — |
case-20 | pass→pass | 8,799 | 12,060 | +37% | 1 | 1 | 0% | 1,857 | 3,185 | +72% | 0 | 0 | — |
case-22 | pass→pass | 15,832 | 17,588 | +11% | 1 | 1 | 0% | 2,698 | 3,747 | +39% | 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 +14 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 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.