Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Verify a code change by running the app, the relevant command, or a focused server flow and reporting concrete evidence.
.claude/skills/oboard-verify/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -10% | 0% |
Use this skill when a task is not finished until the change is exercised.
Produce a short verification result grounded in execution, not inference. Prefer the narrowest check that proves the changed behavior works.
See examples/cli.md.
See examples/server.md.
Verified: what you ran and what passed.Not verified: anything you could not run.Risk: the main remaining uncertainty, if any.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,981 | 3,101 | -56% | 1 | 1 | 0% | 1,141 | 493 | -57% | 0 | 0 | — |
case-02 | fail→fail | 3,093 | 2,421 | -22% | 1 | 1 | 0% | 337 | 496 | +47% | 0 | 0 | — |
case-03 | fail→fail | 7,331 | 3,558 | -51% | 1 | 1 | 0% | 1,094 | 562 | -49% | 0 | 0 | — |
case-04 | fail→pass | 4,483 | 3,243 | -28% | 1 | 1 | 0% | 769 | 913 | +19% | 0 | 0 | — |
case-05 | fail→fail | 9,902 | 8,069 | -19% | 1 | 1 | 0% | 1,712 | 1,629 | -5% | 0 | 0 | — |
case-06 | fail→pass | 10,150 | 3,952 | -61% | 1 | 1 | 0% | 1,673 | 1,084 | -35% | 0 | 0 | — |
case-07 | fail→pass | 6,896 | 5,381 | -22% | 1 | 1 | 0% | 1,086 | 1,279 | +18% | 0 | 0 | — |
case-08 | pass→pass | 9,495 | 4,997 | -47% | 1 | 1 | 0% | 1,692 | 1,170 | -31% | 0 | 0 | — |
case-09 | fail→fail | 8,819 | 5,999 | -32% | 1 | 1 | 0% | 1,376 | 1,390 | +1% | 0 | 0 | — |
case-10 | fail→fail | 8,471 | 3,964 | -53% | 1 | 1 | 0% | 1,363 | 1,033 | -24% | 0 | 0 | — |
case-11 | fail→fail | 7,574 | 4,373 | -42% | 1 | 1 | 0% | 1,225 | 1,053 | -14% | 0 | 0 | — |
case-22 | pass→pass | 21,712 | 16,622 | -23% | 1 | 1 | 0% | 3,705 | 3,190 | -14% | 0 | 0 | — |
case-12 | fail→fail | 5,915 | 3,207 | -46% | 1 | 1 | 0% | 958 | 828 | -14% | 0 | 0 | — |
case-13 | fail→fail | 6,561 | 4,950 | -25% | 1 | 1 | 0% | 1,061 | 1,243 | +17% | 0 | 0 | — |
case-14 | fail→pass | 9,330 | 5,432 | -42% | 1 | 1 | 0% | 1,558 | 1,297 | -17% | 0 | 0 | — |
case-15 | fail→fail | 8,835 | 6,313 | -29% | 1 | 1 | 0% | 1,531 | 1,383 | -10% | 0 | 0 | — |
case-16 | pass→pass | 5,296 | 3,569 | -33% | 1 | 1 | 0% | 832 | 866 | +4% | 0 | 0 | — |
case-17 | pass→pass | 6,697 | 5,175 | -23% | 1 | 1 | 0% | 1,115 | 1,172 | +5% | 0 | 0 | — |
case-18 | fail→pass | 8,043 | 5,601 | -30% | 1 | 1 | 0% | 1,284 | 1,150 | -10% | 0 | 0 | — |
case-19 | pass→pass | 3,676 | 3,199 | -13% | 1 | 1 | 0% | 640 | 905 | +41% | 0 | 0 | — |
case-20 | pass→fail | 19,151 | 2,627 | -86% | 1 | 1 | 0% | 4,109 | 459 | -89% | 0 | 0 | — |
case-21 | pass→fail | 8,261 | 3,464 | -58% | 1 | 1 | 0% | 1,776 | 542 | -69% | 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 17 counted toward the lift figure. The other 5 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 +14 percentage points is the difference between those two pass rates over the 17 comparable cases. 3 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.