Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Restate technical problem as biological question. Translate engineering challenges into nature's language.
.claude/skills/yogsoth-ai-problem-biologization/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✓→✗ | ▼ Worse | 7% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 30% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 139% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | -19% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | -57% | 0% |
Restate technical problem as biological question.
Subagent — spawned via subagent-spawning/spawn-agent skill.
Biologization requires creative reframing of technical problems into biological language, searching for functional equivalents across kingdoms of life. Benefits from dedicated attention to find non-obvious biological framings.
<!-- 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 | 24,842 | 19,028 | -23% | 1 | 1 | 0% | 2,582 | 2,099 | -19% | 0 | 0 | — |
case-02 | fail→fail | 29,771 | 31,887 | +7% | 1 | 1 | 0% | 3,256 | 1,402 | -57% | 0 | 0 | — |
case-03 | fail→fail | 19,537 | 23,928 | +22% | 1 | 1 | 0% | 2,777 | 3,591 | +29% | 0 | 0 | — |
case-04 | pass→pass | 21,522 | 46,288 | +115% | 1 | 1 | 0% | 4,460 | 5,809 | +30% | 0 | 0 | — |
case-05 | pass→pass | 34,870 | 34,582 | -1% | 1 | 1 | 0% | 2,276 | 5,445 | +139% | 0 | 0 | — |
case-06 | pass→fail | 16,784 | 35,836 | +114% | 1 | 1 | 0% | 2,983 | 3,201 | +7% | 0 | 0 | — |
case-07 | fail→fail | 12,811 | 9,316 | -27% | 1 | 1 | 0% | 1,157 | 1,471 | +27% | 0 | 0 | — |
case-08 | fail→fail | 13,977 | 14,916 | +7% | 1 | 1 | 0% | 1,270 | 1,610 | +27% | 0 | 0 | — |
case-09 | fail→fail | 10,397 | 13,564 | +30% | 1 | 1 | 0% | 734 | 1,350 | +84% | 0 | 0 | — |
case-10 | fail→fail | 11,694 | 11,981 | +2% | 1 | 1 | 0% | 938 | 1,076 | +15% | 0 | 0 | — |
case-11 | fail→fail | 25,108 | 22,888 | -9% | 1 | 1 | 0% | 2,892 | 3,050 | +5% | 0 | 0 | — |
case-12 | fail→fail | 14,036 | 13,068 | -7% | 1 | 1 | 0% | 2,012 | 1,901 | -6% | 0 | 0 | — |
case-13 | fail→fail | 21,280 | 19,857 | -7% | 1 | 1 | 0% | 2,653 | 2,539 | -4% | 0 | 0 | — |
case-14 | fail→fail | 13,906 | 23,606 | +70% | 1 | 1 | 0% | 1,246 | 3,140 | +152% | 0 | 0 | — |
case-15 | fail→fail | 11,899 | 24,665 | +107% | 1 | 1 | 0% | 1,337 | 2,589 | +94% | 0 | 0 | — |
case-16 | fail→fail | 13,728 | 13,372 | -3% | 1 | 1 | 0% | 1,115 | 1,401 | +26% | 0 | 0 | — |
case-17 | fail→fail | 23,428 | 21,371 | -9% | 1 | 1 | 0% | 2,924 | 2,632 | -10% | 0 | 0 | — |
case-18 | fail→fail | 13,097 | 9,358 | -29% | 1 | 1 | 0% | 1,049 | 771 | -27% | 0 | 0 | — |
case-19 | fail→fail | 8,642 | 6,541 | -24% | 1 | 1 | 0% | 661 | 1,146 | +73% | 0 | 0 | — |
case-20 | fail→fail | 23,598 | 11,866 | -50% | 1 | 1 | 0% | 2,727 | 1,919 | -30% | 0 | 0 | — |
case-21 | fail→fail | 11,998 | 22,437 | +87% | 1 | 1 | 0% | 1,840 | 2,747 | +49% | 0 | 0 | — |
case-22 | fail→fail | 14,565 | 10,213 | -30% | 1 | 1 | 0% | 1,271 | 1,574 | +24% | 0 | 0 | — |
case-23 | fail→fail | 14,615 | 7,468 | -49% | 1 | 1 | 0% | 1,430 | 1,201 | -16% | 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. 23 cases were attempted, and 22 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 -4 percentage points is the difference between those two pass rates over the 22 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.