Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Compatibility alias — renamed to fitness.
.claude/skills/boshu2-goals/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -93% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -71% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -90% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -82% | 0% |
This skill was renamed to fitness on 2026-07-29. Everything it did lives there unchanged; the ao goals CLI command family keeps its name.
When invoked, apply skills/fitness/SKILL.md exactly as written — this alias adds no behavior, grants no authority, and produces nothing of its own. It exists so existing references and habits keep resolving, and it is deliberately not advertised as a destination.
Physical deletion follows the observed-zero policy: this alias is removed only after a declared observation window shows no remaining use.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 5,701 | 1,612 | -72% | 1 | 1 | 0% | 876 | 364 | -58% | 0 | 0 | — |
case-02 | pass→pass | 10,192 | 2,627 | -74% | 1 | 1 | 0% | 1,525 | 437 | -71% | 0 | 0 | — |
case-03 | fail→pass | 24,539 | 1,409 | -94% | 1 | 1 | 0% | 4,359 | 323 | -93% | 0 | 0 | — |
case-04 | fail→pass | 16,541 | 2,851 | -83% | 1 | 1 | 0% | 2,236 | 645 | -71% | 0 | 0 | — |
case-05 | pass→pass | 11,576 | 2,403 | -79% | 1 | 1 | 0% | 1,699 | 545 | -68% | 0 | 0 | — |
case-06 | fail→pass | 61,286 | 2,733 | -96% | 1 | 1 | 0% | 5,053 | 526 | -90% | 0 | 0 | — |
case-07 | fail→pass | 14,351 | 1,963 | -86% | 1 | 1 | 0% | 2,087 | 375 | -82% | 0 | 0 | — |
case-08 | fail→pass | 8,331 | 4,061 | -51% | 1 | 1 | 0% | 658 | 439 | -33% | 0 | 0 | — |
case-09 | pass→pass | 8,151 | 2,415 | -70% | 1 | 1 | 0% | 1,110 | 561 | -49% | 0 | 0 | — |
case-10 | fail→pass | 13,168 | 1,931 | -85% | 1 | 1 | 0% | 2,139 | 374 | -83% | 0 | 0 | — |
case-11 | fail→pass | 9,130 | 1,789 | -80% | 1 | 1 | 0% | 1,529 | 331 | -78% | 0 | 0 | — |
case-12 | pass→pass | 16,469 | 1,357 | -92% | 1 | 1 | 0% | 2,508 | 300 | -88% | 0 | 0 | — |
case-13 | pass→pass | 6,728 | 2,213 | -67% | 1 | 1 | 0% | 977 | 420 | -57% | 0 | 0 | — |
case-14 | fail→fail | 12,429 | 2,156 | -83% | 1 | 1 | 0% | 2,062 | 387 | -81% | 0 | 0 | — |
case-15 | pass→pass | 17,382 | 2,536 | -85% | 1 | 1 | 0% | 2,944 | 326 | -89% | 0 | 0 | — |
case-16 | pass→pass | 15,761 | 1,426 | -91% | 1 | 1 | 0% | 2,626 | 345 | -87% | 0 | 0 | — |
case-17 | fail→pass | 6,890 | 1,951 | -72% | 1 | 1 | 0% | 952 | 421 | -56% | 0 | 0 | — |
case-18 | pass→pass | 4,572 | 2,308 | -50% | 1 | 1 | 0% | 671 | 470 | -30% | 0 | 0 | — |
case-19 | pass→pass | 12,118 | 2,249 | -81% | 1 | 1 | 0% | 1,691 | 453 | -73% | 0 | 0 | — |
case-20 | pass→pass | 4,354 | 3,847 | -12% | 1 | 1 | 0% | 733 | 815 | +11% | 0 | 0 | — |
case-21 | pass→pass | 3,682 | 2,025 | -45% | 1 | 1 | 0% | 500 | 415 | -17% | 0 | 0 | — |
case-22 | pass→pass | 2,607 | 2,116 | -19% | 1 | 1 | 0% | 270 | 465 | +72% | 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 20 counted toward the lift figure. The other 2 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 +41 percentage points is the difference between those two pass rates over the 20 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.