Install any skill in seconds. Free to start, no credit card required.
Get Started Free →What is the worst case? — Extreme condition construction and failure mode enumeration for risk preparedness
.claude/skills/yogsoth-ai-stress-scenario/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 7% | 0% |
Stress Testing combined with Failure Mode Analysis. Deliberately construct extreme but plausible scenarios that maximally challenge the research approach. Identify breaking points, failure modes, and minimum conditions for viability.
Key principles:
scenario-driver-identificationworst-case-construction (×3-5)scenario-impact-assessment (per worst case)robustness-scoringscenario-synthesis| Step | Token Budget | Notes | |------|-------------|-------| | Driver identification | 8K | Threat-focused | | Worst-case construction | 12K × N | N = 3-5 stress scenarios | | Impact assessment | 12K × N | Deeper analysis for stress | | Robustness scoring | 10K | Failure threshold mapping | | Synthesis | 12K | Risk mitigation report |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | strategy-robustness-testing | Orchestrates impact assessment and robustness scoring to evaluate research approach resilience across scenarios |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | robustness-scoring | Compute robustness index across scenarios with sensitivity analysis | | scenario-driver-identification | Identify key uncertainty drivers using PESTEL framework scanning | | scenario-impact-assessment | Assess each scenario's impact on the research approach across multiple dimensions | | scenario-synthesis | Comprehensive scenario analysis report synthesizing all scenario work | | worst-case-construction | Construct extreme but plausible worst-case scenarios for stress testing |
<!-- 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 | 50,793 | 77,084 | +52% | 1 | 1 | 0% | 6,778 | 5,725 | -16% | 0 | 0 | — |
case-02 | pass→fail | 36,678 | 72,679 | +98% | 1 | 1 | 0% | 4,832 | 2,453 | -49% | 0 | 0 | — |
case-03 | pass→pass | 27,098 | 25,013 | -8% | 1 | 1 | 0% | 3,960 | 3,794 | -4% | 0 | 0 | — |
case-09 | fail→pass | 18,988 | 4,952 | -74% | 1 | 1 | 0% | 2,085 | 1,455 | -30% | 0 | 0 | — |
case-04 | pass→pass | 28,089 | 52,025 | +85% | 1 | 1 | 0% | 3,033 | 8,051 | +165% | 0 | 0 | — |
case-05 | pass→fail | 25,338 | 15,800 | -38% | 1 | 1 | 0% | 2,928 | 1,249 | -57% | 0 | 0 | — |
case-06 | pass→pass | 27,534 | 36,366 | +32% | 1 | 1 | 0% | 3,673 | 5,832 | +59% | 0 | 0 | — |
case-07 | pass→pass | 24,733 | 35,380 | +43% | 1 | 1 | 0% | 2,950 | 5,301 | +80% | 0 | 0 | — |
case-08 | pass→fail | 35,422 | 25,567 | -28% | 1 | 1 | 0% | 5,386 | 4,135 | -23% | 0 | 0 | — |
case-10 | fail→pass | 23,817 | 18,817 | -21% | 1 | 1 | 0% | 2,924 | 3,099 | +6% | 0 | 0 | — |
case-11 | fail→pass | 7,260 | 7,875 | +8% | 1 | 1 | 0% | 1,141 | 1,071 | -6% | 0 | 0 | — |
case-12 | pass→pass | 15,872 | 17,790 | +12% | 1 | 1 | 0% | 2,349 | 2,437 | +4% | 0 | 0 | — |
case-13 | pass→pass | 15,928 | 24,140 | +52% | 1 | 1 | 0% | 1,610 | 3,281 | +104% | 0 | 0 | — |
case-14 | fail→pass | 45,721 | 22,347 | -51% | 1 | 1 | 0% | 3,880 | 4,152 | +7% | 0 | 0 | — |
case-15 | fail→pass | 21,824 | 12,078 | -45% | 1 | 1 | 0% | 2,648 | 1,705 | -36% | 0 | 0 | — |
case-16 | fail→pass | 15,657 | 14,356 | -8% | 1 | 1 | 0% | 1,587 | 2,074 | +31% | 0 | 0 | — |
case-17 | fail→pass | 21,449 | 32,034 | +49% | 1 | 1 | 0% | 2,608 | 4,710 | +81% | 0 | 0 | — |
case-18 | fail→pass | 12,958 | 20,915 | +61% | 1 | 1 | 0% | 1,165 | 2,943 | +153% | 0 | 0 | — |
case-19 | fail→pass | 14,440 | 4,020 | -72% | 1 | 1 | 0% | 2,379 | 1,392 | -41% | 0 | 0 | — |
case-20 | pass→pass | 14,551 | 25,358 | +74% | 1 | 1 | 0% | 2,637 | 4,327 | +64% | 0 | 0 | — |
case-21 | pass→pass | 22,459 | 24,729 | +10% | 1 | 1 | 0% | 3,572 | 5,727 | +60% | 0 | 0 | — |
case-22 | pass→fail | 21,957 | 47,945 | +118% | 1 | 1 | 0% | 2,522 | 7,225 | +186% | 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, and 21 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 +27 percentage points is the difference between those two pass rates over the 21 comparable cases. 4 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.