Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build concept levels from purpose through concepts to ideas, with escape and fractionation at each level.
.claude/skills/yogsoth-ai-concept-hierarchy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 40% | 0% |
Build multi-level concept structures from purpose through broad concepts to specific ideas.
Use concept-fan-expansion SOP to build the initial hierarchy:
Apply escape-technique SOP at each level to identify and break free from dominant thinking patterns that limit the hierarchy.
Use fractionation SOP to split broad concepts into finer units, enabling recombination into novel structures.
| Metric | Floor | |--------|-------| | Concept levels | ≥3 | | Ideas per level | ≥2 | | Escape directions identified | ≥2 | | Fractionation recombinations | ≥3 |
| SOP | Role | |-----|------| | concept-fan-expansion | Stage 1 — build concept hierarchy | | escape-technique | Stage 2 — escape dominant patterns per level | | fractionation | Stage 3 — split and recombine concepts |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | concept-fan-expansion | Expand concept fan from purpose through concepts to directions to ideas (de Bono Concept Fan). | | escape-technique | Identify dominant thinking pattern and escape it via deliberate pattern-breaking. | | fractionation | Split concepts into smaller units and recombine them differently to produce novel structures. |
<!-- 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→pass | 24,012 | 16,217 | -32% | 1 | 1 | 0% | 3,907 | 3,002 | -23% | 0 | 0 | — |
case-02 | fail→pass | 26,569 | 23,957 | -10% | 1 | 1 | 0% | 3,949 | 4,283 | +8% | 0 | 0 | — |
case-03 | fail→pass | 24,259 | 25,285 | +4% | 1 | 1 | 0% | 3,674 | 4,403 | +20% | 0 | 0 | — |
case-04 | fail→pass | 27,872 | 26,758 | -4% | 1 | 1 | 0% | 4,092 | 4,531 | +11% | 0 | 0 | — |
case-05 | fail→pass | 17,876 | 22,526 | +26% | 1 | 1 | 0% | 2,708 | 3,802 | +40% | 0 | 0 | — |
case-06 | fail→pass | 18,888 | 21,309 | +13% | 1 | 1 | 0% | 2,828 | 3,667 | +30% | 0 | 0 | — |
case-07 | fail→pass | 28,756 | 38,105 | +33% | 1 | 1 | 0% | 4,555 | 5,979 | +31% | 0 | 0 | — |
case-08 | pass→pass | 22,529 | 25,001 | +11% | 1 | 1 | 0% | 3,281 | 3,903 | +19% | 0 | 0 | — |
case-09 | pass→pass | 22,582 | 22,128 | -2% | 1 | 1 | 0% | 3,483 | 3,657 | +5% | 0 | 0 | — |
case-10 | pass→pass | 26,177 | 21,172 | -19% | 1 | 1 | 0% | 3,900 | 3,604 | -8% | 0 | 0 | — |
case-11 | fail→pass | 20,797 | 22,157 | +7% | 1 | 1 | 0% | 3,129 | 4,189 | +34% | 0 | 0 | — |
case-12 | pass→pass | 16,337 | 23,638 | +45% | 1 | 1 | 0% | 2,494 | 4,190 | +68% | 0 | 0 | — |
case-13 | fail→pass | 16,753 | 22,340 | +33% | 1 | 1 | 0% | 2,938 | 3,616 | +23% | 0 | 0 | — |
case-14 | fail→pass | 20,593 | 19,820 | -4% | 1 | 1 | 0% | 2,917 | 3,557 | +22% | 0 | 0 | — |
case-15 | pass→pass | 23,212 | 29,619 | +28% | 1 | 1 | 0% | 3,572 | 5,183 | +45% | 0 | 0 | — |
case-16 | pass→pass | 29,471 | 29,630 | +1% | 1 | 1 | 0% | 4,617 | 4,908 | +6% | 0 | 0 | — |
case-17 | fail→pass | 18,394 | 16,857 | -8% | 1 | 1 | 0% | 2,619 | 3,348 | +28% | 0 | 0 | — |
case-18 | pass→pass | 25,257 | 21,391 | -15% | 1 | 1 | 0% | 3,756 | 3,662 | -3% | 0 | 0 | — |
case-19 | pass→pass | 23,352 | 25,980 | +11% | 1 | 1 | 0% | 3,562 | 4,362 | +22% | 0 | 0 | — |
case-20 | fail→pass | 16,669 | 17,616 | +6% | 1 | 1 | 0% | 2,467 | 3,176 | +29% | 0 | 0 | — |
case-21 | pass→pass | 10,968 | 11,379 | +4% | 1 | 1 | 0% | 1,571 | 2,077 | +32% | 0 | 0 | — |
case-22 | pass→pass | 15,125 | 20,547 | +36% | 1 | 1 | 0% | 2,824 | 4,129 | +46% | 0 | 0 | — |
case-23 | pass→pass | 21,358 | 16,500 | -23% | 1 | 1 | 0% | 3,360 | 3,013 | -10% | 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. The headline lift of +52 percentage points is the difference between those two pass rates over the 23 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.