Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Construct distinct future scenarios spanning key uncertainties for portfolio stress testing.
.claude/skills/yogsoth-ai-scenario-construction/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 85% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 167% | 0% |
Build a set of distinct, plausible future scenarios that span the key uncertainties relevant to portfolio performance.
Spawns a subagent that identifies key uncertainties and constructs diverse scenarios for stress testing.
Scenario construction requires creative yet disciplined thinking to produce scenarios that are distinct, plausible, and collectively span the uncertainty space. This generative-analytical work benefits from focused attention.
Output must contain at least 3 distinct scenarios that span different combinations of key uncertainties. Each scenario must include a narrative, key assumptions, and probability estimate.
<!-- 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→pass | 13,901 | 36,742 | +164% | 1 | 1 | 0% | 2,322 | 4,288 | +85% | 0 | 0 | — |
case-02 | fail→pass | 24,063 | 31,989 | +33% | 1 | 1 | 0% | 3,710 | 5,398 | +45% | 0 | 0 | — |
case-03 | fail→pass | 22,984 | 22,100 | -4% | 1 | 1 | 0% | 2,516 | 3,553 | +41% | 0 | 0 | — |
case-04 | fail→pass | 23,388 | 37,004 | +58% | 1 | 1 | 0% | 2,416 | 4,927 | +104% | 0 | 0 | — |
case-05 | fail→pass | 17,323 | 27,833 | +61% | 1 | 1 | 0% | 1,765 | 4,708 | +167% | 0 | 0 | — |
case-06 | fail→fail | 31,951 | 39,415 | +23% | 1 | 1 | 0% | 4,693 | 5,311 | +13% | 0 | 0 | — |
case-07 | fail→pass | 15,850 | 25,384 | +60% | 1 | 1 | 0% | 2,083 | 4,096 | +97% | 0 | 0 | — |
case-08 | fail→pass | 21,717 | 35,108 | +62% | 1 | 1 | 0% | 3,517 | 6,429 | +83% | 0 | 0 | — |
case-09 | fail→pass | 18,528 | 27,484 | +48% | 1 | 1 | 0% | 2,481 | 4,257 | +72% | 0 | 0 | — |
case-10 | fail→pass | 21,911 | 37,123 | +69% | 1 | 1 | 0% | 3,363 | 5,177 | +54% | 0 | 0 | — |
case-11 | fail→pass | 25,550 | 37,132 | +45% | 1 | 1 | 0% | 3,220 | 4,999 | +55% | 0 | 0 | — |
case-12 | fail→pass | 14,578 | 21,464 | +47% | 1 | 1 | 0% | 1,984 | 2,988 | +51% | 0 | 0 | — |
case-13 | fail→pass | 24,696 | 43,591 | +77% | 1 | 1 | 0% | 2,963 | 6,796 | +129% | 0 | 0 | — |
case-14 | fail→pass | 11,053 | 35,968 | +225% | 1 | 1 | 0% | 2,309 | 6,071 | +163% | 0 | 0 | — |
case-15 | pass→pass | 24,886 | 29,640 | +19% | 1 | 1 | 0% | 3,739 | 5,260 | +41% | 0 | 0 | — |
case-16 | fail→pass | 24,994 | 38,314 | +53% | 1 | 1 | 0% | 3,035 | 5,735 | +89% | 0 | 0 | — |
case-17 | fail→pass | 19,713 | 40,716 | +107% | 1 | 1 | 0% | 2,555 | 6,589 | +158% | 0 | 0 | — |
case-18 | fail→pass | 15,456 | 51,690 | +234% | 1 | 1 | 0% | 2,186 | 5,906 | +170% | 0 | 0 | — |
case-19 | fail→fail | 28,172 | 54,747 | +94% | 1 | 1 | 0% | 3,322 | 7,548 | +127% | 0 | 0 | — |
case-20 | pass→pass | 15,643 | 19,636 | +26% | 1 | 1 | 0% | 2,082 | 3,945 | +89% | 0 | 0 | — |
case-21 | pass→pass | 16,930 | 34,761 | +105% | 1 | 1 | 0% | 2,911 | 6,810 | +134% | 0 | 0 | — |
case-22 | pass→pass | 16,562 | 41,112 | +148% | 1 | 1 | 0% | 2,282 | 7,098 | +211% | 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. The headline lift of +73 percentage points is the difference between those two pass rates over the 22 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.