Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Role-play as reviewer/practitioner/theorist/novice/competitor to generate diverse perspectives on a solution.
.claude/skills/yogsoth-ai-role-based-ideation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -51% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -20% | 0% |
Role-play as reviewer/practitioner/theorist/novice/competitor to generate diverse perspectives on a solution.
| Resource | Target | Current | % | |----------|--------|---------|---| | web-search | 20 | 0 | 0% | | web-research | 5 | 0 | 0% | | paper-overview | 20 | 0 | 0% | | paper-search | 12 | 0 | 0% | | paper-research | 5 | 0 | 0% |
Cannot exit strategy until ≥80% of each budget line is consumed OR yield targets are met with justification for remaining budget.
| Tactic | Role | |--------|------| | perspective-rotation | Rotate through all five role perspectives systematically | | evaluation-filtering | Filter and rank resulting ideas (shared) |
| SOP | Role | |-----|------| | reviewer2-hat | Hostile reviewer perspective | | practitioner-hat | Engineer perspective | | theorist-hat | Theorist perspective | | novice-perspective | Novice perspective | | competitor-simulation | Competitor perspective | | perspective-synthesis | Synthesize all role outputs |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | competitor-simulation | Competitor perspective — design strategies to defeat this solution, then use attack vectors to improve it. | | novice-perspective | Novice perspective — question the 'obvious' by adopting deliberate ignorance to reveal hidden complexity. | | perspective-synthesis | Synthesize all perspective outputs into a structured multi-perspective idea report. | | practitioner-hat | Engineer perspective — assess buildability, cost, timeline, and integration challenges. | | reviewer2-hat | Hostile reviewer perspective — find fatal flaws, logical gaps, and missing evidence in a solution. | | theorist-hat | Theorist perspective — assess theoretical foundations, formal rigor, and formalization opportunities. |
<!-- 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 | pass→fail | 55,553 | 22,065 | -60% | 1 | 1 | 0% | 7,607 | 1,268 | -83% | 0 | 0 | — |
case-23 | pass→pass | 7,459 | 10,217 | +37% | 1 | 1 | 0% | 388 | 1,443 | +272% | 0 | 0 | — |
case-02 | fail→fail | 27,331 | 22,950 | -16% | 1 | 1 | 0% | 3,234 | 1,394 | -57% | 0 | 0 | — |
case-03 | fail→fail | 15,064 | 21,253 | +41% | 1 | 1 | 0% | 1,431 | 1,259 | -12% | 0 | 0 | — |
case-04 | pass→pass | 32,582 | 65,529 | +101% | 1 | 1 | 0% | 3,806 | 8,218 | +116% | 0 | 0 | — |
case-05 | pass→pass | 14,363 | 66,577 | +364% | 1 | 1 | 0% | 2,062 | 8,822 | +328% | 0 | 0 | — |
case-06 | pass→fail | 29,699 | 19,748 | -34% | 1 | 1 | 0% | 3,587 | 1,254 | -65% | 0 | 0 | — |
case-07 | pass→fail | 18,548 | 67,703 | +265% | 1 | 1 | 0% | 1,875 | 1,157 | -38% | 0 | 0 | — |
case-08 | pass→pass | 22,873 | 37,798 | +65% | 1 | 1 | 0% | 2,670 | 5,325 | +99% | 0 | 0 | — |
case-09 | fail→pass | 20,633 | 8,298 | -60% | 1 | 1 | 0% | 2,345 | 1,143 | -51% | 0 | 0 | — |
case-10 | fail→pass | 16,282 | 8,774 | -46% | 1 | 1 | 0% | 2,818 | 1,332 | -53% | 0 | 0 | — |
case-11 | fail→pass | 33,106 | 3,151 | -90% | 1 | 1 | 0% | 2,023 | 1,210 | -40% | 0 | 0 | — |
case-12 | fail→pass | 20,564 | 12,740 | -38% | 1 | 1 | 0% | 2,538 | 1,807 | -29% | 0 | 0 | — |
case-13 | fail→pass | 15,839 | 8,735 | -45% | 1 | 1 | 0% | 1,518 | 1,215 | -20% | 0 | 0 | — |
case-14 | fail→pass | 17,173 | 8,214 | -52% | 1 | 1 | 0% | 1,877 | 1,130 | -40% | 0 | 0 | — |
case-15 | pass→pass | 13,918 | 5,657 | -59% | 1 | 1 | 0% | 1,291 | 1,631 | +26% | 0 | 0 | — |
case-16 | pass→fail | 24,923 | 20,061 | -20% | 1 | 1 | 0% | 3,039 | 1,252 | -59% | 0 | 0 | — |
case-17 | fail→fail | 23,454 | 15,952 | -32% | 1 | 1 | 0% | 2,879 | 1,254 | -56% | 0 | 0 | — |
case-18 | pass→pass | 20,141 | 35,069 | +74% | 1 | 1 | 0% | 2,483 | 6,186 | +149% | 0 | 0 | — |
case-19 | pass→fail | 18,602 | 18,596 | -0% | 1 | 1 | 0% | 2,029 | 1,010 | -50% | 0 | 0 | — |
case-20 | fail→pass | 17,281 | 33,818 | +96% | 1 | 1 | 0% | 1,688 | 3,043 | +80% | 0 | 0 | — |
case-21 | pass→pass | 11,072 | 16,636 | +50% | 1 | 1 | 0% | 1,048 | 2,623 | +150% | 0 | 0 | — |
case-22 | pass→fail | 7,537 | 30,227 | +301% | 1 | 1 | 0% | 1,353 | 5,068 | +275% | 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 15 counted toward the lift figure. The other 8 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 15 comparable cases. 7 cases got worse with the skill loaded, and they are included in that figure.
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.