Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Pair programming continuation
.claude/skills/sterlingcrispin-pair/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-16 | ✗→✓ | ▲ Improved | 811% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -89% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -2% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -10% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -12% | 0% |
continue pair programming:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-07 | fail→fail | 4,132 | 4,540 | +10% | 1 | 1 | 0% | 240 | 275 | +15% | 0 | 0 | — |
case-01 | fail→fail | 42,902 | 9,578 | -78% | 1 | 1 | 0% | 296 | 232 | -22% | 0 | 0 | — |
case-02 | fail→fail | 3,922 | 3,825 | -2% | 1 | 1 | 0% | 175 | 154 | -12% | 0 | 0 | — |
case-03 | pass→pass | 9,100 | 8,415 | -8% | 1 | 1 | 0% | 1,757 | 1,723 | -2% | 0 | 0 | — |
case-04 | pass→pass | 12,052 | 10,588 | -12% | 1 | 1 | 0% | 2,171 | 1,949 | -10% | 0 | 0 | — |
case-05 | pass→pass | 10,591 | 9,007 | -15% | 1 | 1 | 0% | 1,958 | 1,732 | -12% | 0 | 0 | — |
case-06 | fail→fail | 7,893 | 10,856 | +38% | 1 | 1 | 0% | 1,302 | 1,465 | +13% | 0 | 0 | — |
case-08 | fail→fail | 27,859 | 3,955 | -86% | 1 | 1 | 0% | 201 | 236 | +17% | 0 | 0 | — |
case-09 | fail→fail | 4,595 | 14,864 | +223% | 1 | 1 | 0% | 274 | 240 | -12% | 0 | 0 | — |
case-10 | fail→fail | 4,519 | 4,426 | -2% | 1 | 1 | 0% | 250 | 271 | +8% | 0 | 0 | — |
case-11 | fail→fail | 4,019 | 4,402 | +10% | 1 | 1 | 0% | 242 | 237 | -2% | 0 | 0 | — |
case-12 | fail→fail | 3,519 | 4,495 | +28% | 1 | 1 | 0% | 201 | 265 | +32% | 0 | 0 | — |
case-13 | fail→fail | 2,614 | 18,352 | +602% | 1 | 1 | 0% | 177 | 166 | -6% | 0 | 0 | — |
case-14 | fail→fail | 3,349 | 3,262 | -3% | 1 | 1 | 0% | 223 | 188 | -16% | 0 | 0 | — |
case-15 | fail→fail | 4,521 | 4,163 | -8% | 1 | 1 | 0% | 243 | 188 | -23% | 0 | 0 | — |
case-16 | fail→pass | 44,642 | 20,598 | -54% | 1 | 1 | 0% | 238 | 2,168 | +811% | 0 | 0 | — |
case-17 | fail→fail | 3,993 | 3,497 | -12% | 1 | 1 | 0% | 205 | 140 | -32% | 0 | 0 | — |
case-18 | fail→fail | 4,280 | 4,469 | +4% | 1 | 1 | 0% | 216 | 243 | +13% | 0 | 0 | — |
case-19 | fail→fail | 15,617 | 4,162 | -73% | 1 | 1 | 0% | 2,626 | 213 | -92% | 0 | 0 | — |
case-20 | pass→fail | 10,536 | 4,278 | -59% | 1 | 1 | 0% | 1,935 | 214 | -89% | 0 | 0 | — |
case-21 | fail→fail | 5,466 | 27,506 | +403% | 1 | 1 | 0% | 385 | 200 | -48% | 0 | 0 | — |
case-22 | fail→fail | 2,590 | 3,343 | +29% | 1 | 1 | 0% | 210 | 179 | -15% | 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 4 counted toward the lift figure. The other 18 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 -100 percentage points is the difference between those two pass rates over the 4 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
Other measured skills in the registry, with their headline benchmark lift.