Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Apply general knowledge reasoning across diverse domains — trivia, commonsense inference, analogy, and factual question answering.
.claude/skills/a5c-ai-general-knowledge-reasoning/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-21 | ✓→✗ | ▼ Worse | -68% | 0% |
| case-19 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | -12% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | 14% | 0% |
| case-03 | ✗→✗ | ＝ Same ✗ | 7% | 0% |
> Stub — implementation pending.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,382 | 4,669 | -13% | 1 | 1 | 0% | 962 | 849 | -12% | 0 | 0 | — |
case-02 | fail→fail | 5,736 | 5,826 | +2% | 1 | 1 | 0% | 971 | 1,108 | +14% | 0 | 0 | — |
case-03 | fail→fail | 8,087 | 9,161 | +13% | 1 | 1 | 0% | 1,314 | 1,404 | +7% | 0 | 0 | — |
case-04 | fail→fail | 6,081 | 4,779 | -21% | 1 | 1 | 0% | 1,066 | 863 | -19% | 0 | 0 | — |
case-05 | fail→fail | 4,900 | 2,507 | -49% | 1 | 1 | 0% | 817 | 478 | -41% | 0 | 0 | — |
case-06 | fail→fail | 4,576 | 3,167 | -31% | 1 | 1 | 0% | 783 | 565 | -28% | 0 | 0 | — |
case-07 | fail→fail | 5,494 | 2,918 | -47% | 1 | 1 | 0% | 854 | 574 | -33% | 0 | 0 | — |
case-08 | fail→fail | 3,623 | 2,357 | -35% | 1 | 1 | 0% | 647 | 411 | -36% | 0 | 0 | — |
case-09 | fail→fail | 5,723 | 5,349 | -7% | 1 | 1 | 0% | 985 | 941 | -4% | 0 | 0 | — |
case-10 | fail→fail | 4,896 | 3,509 | -28% | 1 | 1 | 0% | 1,004 | 694 | -31% | 0 | 0 | — |
case-11 | fail→fail | 8,287 | 6,728 | -19% | 1 | 1 | 0% | 1,237 | 1,113 | -10% | 0 | 0 | — |
case-12 | fail→fail | 6,957 | 6,695 | -4% | 1 | 1 | 0% | 1,219 | 1,159 | -5% | 0 | 0 | — |
case-13 | fail→fail | 9,096 | 3,940 | -57% | 1 | 1 | 0% | 1,520 | 718 | -53% | 0 | 0 | — |
case-14 | fail→fail | 5,666 | 4,029 | -29% | 1 | 1 | 0% | 1,025 | 713 | -30% | 0 | 0 | — |
case-15 | fail→fail | 3,494 | 2,836 | -19% | 1 | 1 | 0% | 622 | 556 | -11% | 0 | 0 | — |
case-16 | fail→fail | 5,009 | 3,337 | -33% | 1 | 1 | 0% | 954 | 653 | -32% | 0 | 0 | — |
case-17 | fail→fail | 7,774 | 4,908 | -37% | 1 | 1 | 0% | 1,362 | 753 | -45% | 0 | 0 | — |
case-18 | fail→fail | 5,604 | 4,519 | -19% | 1 | 1 | 0% | 894 | 802 | -10% | 0 | 0 | — |
case-19 | pass→pass | 9,578 | 7,563 | -21% | 1 | 1 | 0% | 1,566 | 1,321 | -16% | 0 | 0 | — |
case-20 | fail→fail | 1,343 | 1,280 | -5% | 1 | 1 | 0% | 128 | 147 | +15% | 0 | 0 | — |
case-21 | pass→fail | 6,690 | 2,374 | -65% | 1 | 1 | 0% | 1,210 | 388 | -68% | 0 | 0 | — |
case-22 | fail→fail | 1,568 | 1,438 | -8% | 1 | 1 | 0% | 195 | 200 | +3% | 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 -100 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.