Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generates structured usability test scenarios with realistic tasks, success criteria, and facilitation notes — ready to run with real participants or in a moderated session.
.claude/skills/owl-listener-test-scenario/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 1% | 0% |
You are an expert in writing usability test scenarios that reveal genuine user behavior.
You write test scenarios with realistic tasks, clear success criteria, and structured observation guides.
Brief, realistic backstory that gives the participant a reason to act without leading them.
Specific goal to accomplish. Action-oriented, not question-based. Avoids UI terminology that hints at the answer.
What to watch for: hesitations, facial expressions, verbal comments, navigation choices, error recovery behavior.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 12,843 | 9,277 | -28% | 1 | 1 | 0% | 2,330 | 1,938 | -17% | 0 | 0 | — |
case-02 | fail→pass | 11,402 | 10,462 | -8% | 1 | 1 | 0% | 1,827 | 2,020 | +11% | 0 | 0 | — |
case-03 | fail→fail | 13,441 | 12,601 | -6% | 1 | 1 | 0% | 2,351 | 2,420 | +3% | 0 | 0 | — |
case-04 | pass→pass | 13,581 | 11,246 | -17% | 1 | 1 | 0% | 2,543 | 2,275 | -11% | 0 | 0 | — |
case-05 | pass→pass | 18,123 | 18,794 | +4% | 1 | 1 | 0% | 2,868 | 3,180 | +11% | 0 | 0 | — |
case-06 | pass→pass | 11,260 | 10,120 | -10% | 1 | 1 | 0% | 1,927 | 2,089 | +8% | 0 | 0 | — |
case-07 | fail→pass | 12,618 | 11,624 | -8% | 1 | 1 | 0% | 2,087 | 2,228 | +7% | 0 | 0 | — |
case-08 | fail→pass | 13,409 | 9,729 | -27% | 1 | 1 | 0% | 2,290 | 1,994 | -13% | 0 | 0 | — |
case-09 | fail→fail | 11,450 | 9,712 | -15% | 1 | 1 | 0% | 1,972 | 1,990 | +1% | 0 | 0 | — |
case-10 | fail→pass | 12,605 | 10,839 | -14% | 1 | 1 | 0% | 2,101 | 2,132 | +1% | 0 | 0 | — |
case-11 | pass→pass | 12,788 | 11,063 | -13% | 1 | 1 | 0% | 2,202 | 2,326 | +6% | 0 | 0 | — |
case-12 | fail→pass | 12,961 | 10,698 | -17% | 1 | 1 | 0% | 2,421 | 2,240 | -7% | 0 | 0 | — |
case-13 | fail→pass | 13,541 | 9,111 | -33% | 1 | 1 | 0% | 2,170 | 1,784 | -18% | 0 | 0 | — |
case-14 | pass→pass | 14,980 | 11,944 | -20% | 1 | 1 | 0% | 2,396 | 2,484 | +4% | 0 | 0 | — |
case-15 | fail→pass | 12,066 | 10,618 | -12% | 1 | 1 | 0% | 2,015 | 2,127 | +6% | 0 | 0 | — |
case-16 | fail→pass | 21,883 | 18,154 | -17% | 1 | 1 | 0% | 3,755 | 3,310 | -12% | 0 | 0 | — |
case-17 | fail→pass | 10,672 | 9,627 | -10% | 1 | 1 | 0% | 2,195 | 1,963 | -11% | 0 | 0 | — |
case-18 | fail→fail | 11,621 | 12,193 | +5% | 1 | 1 | 0% | 1,926 | 2,111 | +10% | 0 | 0 | — |
case-19 | fail→pass | 11,177 | 11,187 | +0% | 1 | 1 | 0% | 1,924 | 2,230 | +16% | 0 | 0 | — |
case-20 | fail→pass | 15,525 | 11,693 | -25% | 1 | 1 | 0% | 2,800 | 2,469 | -12% | 0 | 0 | — |
case-21 | pass→pass | 12,915 | 8,809 | -32% | 1 | 1 | 0% | 2,103 | 1,891 | -10% | 0 | 0 | — |
case-22 | fail→pass | 12,708 | 11,254 | -11% | 1 | 1 | 0% | 2,043 | 2,199 | +8% | 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.