Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Determine when additional ideation yields diminishing returns. Analyzes latest idea batch against existing corpus to judge continue/near-saturation/saturated.
.claude/skills/yogsoth-ai-creative-ideation-saturation-detection/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✓→✗ | ▼ Worse | -28% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -22% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | 56% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | 54% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | 262% | 0% |
Determine when to stop generating — diminishing returns analysis.
Subagent — spawned via subagent-spawning/spawn-agent skill.
Saturation judgment requires comparing the entire existing idea corpus against the latest batch. The comparison is context-intensive and benefits from dedicated processing.
<!-- 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 | 10,013 | 14,619 | +46% | 1 | 1 | 0% | 1,635 | 2,521 | +54% | 0 | 0 | — |
case-02 | fail→fail | 3,569 | 10,439 | +192% | 1 | 1 | 0% | 567 | 2,055 | +262% | 0 | 0 | — |
case-03 | fail→fail | 6,594 | 18,914 | +187% | 1 | 1 | 0% | 1,075 | 1,410 | +31% | 0 | 0 | — |
case-04 | pass→pass | 13,161 | 9,352 | -29% | 1 | 1 | 0% | 1,996 | 1,550 | -22% | 0 | 0 | — |
case-05 | pass→fail | 6,832 | 3,999 | -41% | 1 | 1 | 0% | 1,143 | 825 | -28% | 0 | 0 | — |
case-06 | fail→fail | 6,347 | 19,868 | +213% | 1 | 1 | 0% | 1,016 | 3,693 | +263% | 0 | 0 | — |
case-07 | pass→pass | 17,478 | 25,120 | +44% | 1 | 1 | 0% | 2,863 | 4,461 | +56% | 0 | 0 | — |
case-08 | fail→fail | 4,274 | 13,341 | +212% | 1 | 1 | 0% | 659 | 2,541 | +286% | 0 | 0 | — |
case-09 | fail→fail | 11,180 | 23,299 | +108% | 1 | 1 | 0% | 1,744 | 4,028 | +131% | 0 | 0 | — |
case-10 | fail→fail | 5,500 | 17,074 | +210% | 1 | 1 | 0% | 948 | 2,940 | +210% | 0 | 0 | — |
case-11 | fail→fail | 8,338 | 3,854 | -54% | 1 | 1 | 0% | 963 | 744 | -23% | 0 | 0 | — |
case-12 | fail→fail | 8,331 | 22,793 | +174% | 1 | 1 | 0% | 1,317 | 3,361 | +155% | 0 | 0 | — |
case-13 | fail→fail | 11,666 | 30,505 | +161% | 1 | 1 | 0% | 1,928 | 5,306 | +175% | 0 | 0 | — |
case-14 | fail→fail | 9,186 | 8,873 | -3% | 1 | 1 | 0% | 1,503 | 748 | -50% | 0 | 0 | — |
case-15 | fail→fail | 5,894 | 20,187 | +243% | 1 | 1 | 0% | 949 | 3,384 | +257% | 0 | 0 | — |
case-21 | fail→fail | 6,743 | 8,620 | +28% | 1 | 1 | 0% | 1,056 | 504 | -52% | 0 | 0 | — |
case-16 | fail→fail | 8,639 | 21,140 | +145% | 1 | 1 | 0% | 1,459 | 2,952 | +102% | 0 | 0 | — |
case-17 | fail→fail | 17,701 | 7,039 | -60% | 1 | 1 | 0% | 3,100 | 645 | -79% | 0 | 0 | — |
case-18 | fail→fail | 6,260 | 39,698 | +534% | 1 | 1 | 0% | 996 | 6,493 | +552% | 0 | 0 | — |
case-19 | fail→fail | 12,691 | 15,144 | +19% | 1 | 1 | 0% | 2,029 | 2,678 | +32% | 0 | 0 | — |
case-20 | fail→fail | 17,185 | 34,572 | +101% | 1 | 1 | 0% | 2,676 | 6,123 | +129% | 0 | 0 | — |
case-22 | fail→fail | 5,520 | 49,089 | +789% | 1 | 1 | 0% | 951 | 5,361 | +464% | 0 | 0 | — |
case-23 | fail→fail | 7,235 | 55,607 | +669% | 1 | 1 | 0% | 1,161 | 7,092 | +511% | 0 | 0 | — |
case-24 | fail→fail | 6,805 | 26,656 | +292% | 1 | 1 | 0% | 1,066 | 3,588 | +237% | 0 | 0 | — |
case-25 | fail→fail | 7,107 | 25,624 | +261% | 1 | 1 | 0% | 1,036 | 2,913 | +181% | 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. 25 cases were attempted, and 21 counted toward the lift figure. The other 4 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 -33 percentage points is the difference between those two pass rates over the 21 comparable cases. 2 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.