Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Identify non-additive properties from combinations
.claude/skills/yogsoth-ai-emergent-property-identification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✓→✓ | ＝ Same ✓ | 51% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | 22% | 0% |
| case-22 | ✓→✓ | ＝ Same ✓ | 402% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | 48% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | 118% | 0% |
Identify non-additive properties from combinations — properties that exist in the combination but not in any individual component.
Subagent — spawned via subagent-spawning/spawn-agent skill.
Emergent property identification requires careful comparison between predicted additive properties and actual combination properties, plus creative imagination to spot non-obvious emergence. Benefits from dedicated analytical-creative attention.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | spawn-agent | Spawn a customized CC subagent with full MCP tool access. Used by SOPs that declare execution: subagent. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,984 | 27,048 | +35% | 1 | 1 | 0% | 2,986 | 4,422 | +48% | 0 | 0 | — |
case-02 | fail→fail | 20,134 | 38,525 | +91% | 1 | 1 | 0% | 2,958 | 6,460 | +118% | 0 | 0 | — |
case-03 | fail→fail | 16,967 | 28,413 | +67% | 1 | 1 | 0% | 2,892 | 3,663 | +27% | 0 | 0 | — |
case-04 | fail→fail | 24,317 | 39,698 | +63% | 1 | 1 | 0% | 3,695 | 5,119 | +39% | 0 | 0 | — |
case-05 | fail→fail | 20,420 | 35,257 | +73% | 1 | 1 | 0% | 3,375 | 5,852 | +73% | 0 | 0 | — |
case-06 | fail→fail | 19,772 | 29,720 | +50% | 1 | 1 | 0% | 3,184 | 4,608 | +45% | 0 | 0 | — |
case-07 | fail→fail | 22,617 | 20,212 | -11% | 1 | 1 | 0% | 3,095 | 3,401 | +10% | 0 | 0 | — |
case-08 | fail→fail | 20,080 | 29,313 | +46% | 1 | 1 | 0% | 3,397 | 3,697 | +9% | 0 | 0 | — |
case-09 | fail→fail | 14,655 | 25,058 | +71% | 1 | 1 | 0% | 2,403 | 4,964 | +107% | 0 | 0 | — |
case-10 | fail→fail | 19,089 | 26,244 | +37% | 1 | 1 | 0% | 2,734 | 4,610 | +69% | 0 | 0 | — |
case-11 | fail→fail | 17,153 | 20,884 | +22% | 1 | 1 | 0% | 2,625 | 3,241 | +23% | 0 | 0 | — |
case-12 | fail→fail | 17,906 | 23,197 | +30% | 1 | 1 | 0% | 2,758 | 2,010 | -27% | 0 | 0 | — |
case-13 | fail→fail | 19,654 | 25,546 | +30% | 1 | 1 | 0% | 2,966 | 4,093 | +38% | 0 | 0 | — |
case-14 | fail→fail | 20,072 | 21,769 | +8% | 1 | 1 | 0% | 3,500 | 3,721 | +6% | 0 | 0 | — |
case-15 | fail→fail | 15,474 | 24,378 | +58% | 1 | 1 | 0% | 2,842 | 4,230 | +49% | 0 | 0 | — |
case-16 | fail→fail | 22,227 | 36,058 | +62% | 1 | 1 | 0% | 3,412 | 5,014 | +47% | 0 | 0 | — |
case-17 | fail→fail | 15,807 | 32,353 | +105% | 1 | 1 | 0% | 2,395 | 4,748 | +98% | 0 | 0 | — |
case-18 | fail→fail | 18,084 | 20,487 | +13% | 1 | 1 | 0% | 3,043 | 3,492 | +15% | 0 | 0 | — |
case-19 | fail→fail | 17,999 | 21,171 | +18% | 1 | 1 | 0% | 2,743 | 3,383 | +23% | 0 | 0 | — |
case-20 | pass→pass | 8,383 | 13,502 | +61% | 1 | 1 | 0% | 2,033 | 3,066 | +51% | 0 | 0 | — |
case-21 | pass→pass | 18,205 | 21,160 | +16% | 1 | 1 | 0% | 3,817 | 4,655 | +22% | 0 | 0 | — |
case-22 | pass→pass | 2,361 | 11,439 | +384% | 1 | 1 | 0% | 449 | 2,254 | +402% | 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 21 counted toward the lift figure. The other 1 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 0 percentage points is the difference between those two pass rates over the 21 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.