Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Strategy: Research design-level FMEA — function analysis, failure mode identification, severity/occurrence/detection scoring per AIAG-VDA 2019.
.claude/skills/yogsoth-ai-design-fmea/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 92% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-03 | ✓→✗ | ▼ Worse | 94% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 108% | 0% |
| case-13 | ✓→✓ | ＝ Same ✓ | 12% | 0% |
Systematic failure analysis at the research design level: what functions can fail, how, and with what consequence?
| Parameter | S | M | L | |---|---|---|---| | Functions analyzed | 4 | 10 | 20 | | Failure modes per function | 2 | 3 | 5 | | Chain depth (cause levels) | 2 | 4 | 6 | | Mitigations designed | 3 | 8 | 15 |
function-analysis → failure-mode-extraction → failure-chain-construction
→ [severity-scoring, occurrence-scoring, detection-scoring] (parallel)
→ action-priority-matrix → mitigation-design-sop
→ re-scoring (mitigation-validation tactic)<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | failure-chain-tracing | Tactic: Trace upstream causes and downstream effects of each failure mode. Builds multi-level cause-mode-effect chains for systemic understanding. | | mitigation-validation | Tactic: Run mini-FMEA on proposed mitigations to verify they do not introduce new failure modes. Prevents mitigation-induced risks. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | action-priority-matrix | Compute Risk Priority Number (RPN = S x O x D), classify failure modes into H/M/L action priority per AIAG-VDA tables. | | detection-scoring | Rate detectability 1-10 (inverted: 10 = hardest to detect). Estimates how likely current controls would catch the failure before impact. | | failure-chain-construction | Build cause-mode-effect chains tracing upstream root causes and downstream cascading effects for each failure mode. | | failure-mode-extraction | Extract structured failure mode list from raw scenarios or artifact analysis. Produces standardized failure mode records. | | function-analysis | FMEA Step 3: Decompose artifact into function tree — identify what each component is supposed to do before analyzing how it can fail. | | mitigation-design-sop | Design prevention, detection, and response measures for high-priority failure modes. Produces actionable countermeasure specifications. | | occurrence-scoring | Rate failure mode occurrence probability 1-10. Estimates how likely each failure mode is to manifest during research execution. | | severity-scoring | Rate failure mode severity 1-10 based on end-effect impact. Follows AIAG-VDA severity scale calibrated for research artifacts. |
<!-- 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-13 | pass→pass | 23,736 | 22,687 | -4% | 1 | 1 | 0% | 4,035 | 4,517 | +12% | 0 | 0 | — |
case-14 | pass→pass | 14,488 | 9,267 | -36% | 1 | 1 | 0% | 2,161 | 2,306 | +7% | 0 | 0 | — |
case-11 | pass→pass | 23,470 | 25,834 | +10% | 1 | 1 | 0% | 3,611 | 4,931 | +37% | 0 | 0 | — |
case-01 | fail→pass | 22,592 | 54,327 | +140% | 1 | 1 | 0% | 3,664 | 7,036 | +92% | 0 | 0 | — |
case-02 | fail→pass | 32,198 | 37,848 | +18% | 1 | 1 | 0% | 5,186 | 7,027 | +35% | 0 | 0 | — |
case-03 | pass→fail | 18,966 | 53,329 | +181% | 1 | 1 | 0% | 3,103 | 6,030 | +94% | 0 | 0 | — |
case-04 | pass→pass | 20,077 | 22,001 | +10% | 1 | 1 | 0% | 2,970 | 4,201 | +41% | 0 | 0 | — |
case-12 | pass→pass | 19,378 | 22,313 | +15% | 1 | 1 | 0% | 2,831 | 4,386 | +55% | 0 | 0 | — |
case-05 | pass→fail | 20,671 | 37,254 | +80% | 1 | 1 | 0% | 3,359 | 6,997 | +108% | 0 | 0 | — |
case-06 | pass→pass | 15,657 | 29,216 | +87% | 1 | 1 | 0% | 2,236 | 5,619 | +151% | 0 | 0 | — |
case-07 | pass→pass | 21,504 | 36,115 | +68% | 1 | 1 | 0% | 3,345 | 7,014 | +110% | 0 | 0 | — |
case-08 | pass→pass | 16,940 | 23,989 | +42% | 1 | 1 | 0% | 2,679 | 4,892 | +83% | 0 | 0 | — |
case-09 | fail→fail | 37,726 | 39,398 | +4% | 1 | 1 | 0% | 6,187 | 6,995 | +13% | 0 | 0 | — |
case-10 | pass→pass | 16,912 | 19,745 | +17% | 1 | 1 | 0% | 2,811 | 4,010 | +43% | 0 | 0 | — |
case-15 | pass→pass | 10,655 | 11,897 | +12% | 1 | 1 | 0% | 1,583 | 2,641 | +67% | 0 | 0 | — |
case-16 | pass→pass | 61,427 | 40,189 | -35% | 1 | 1 | 0% | 3,404 | 6,982 | +105% | 0 | 0 | — |
case-17 | pass→pass | 16,095 | 17,421 | +8% | 1 | 1 | 0% | 2,457 | 3,563 | +45% | 0 | 0 | — |
case-18 | pass→pass | 10,481 | 10,636 | +1% | 1 | 1 | 0% | 1,676 | 2,599 | +55% | 0 | 0 | — |
case-19 | pass→pass | 13,577 | 13,025 | -4% | 1 | 1 | 0% | 2,012 | 2,678 | +33% | 0 | 0 | — |
case-20 | pass→pass | 14,045 | 17,150 | +22% | 1 | 1 | 0% | 2,306 | 3,642 | +58% | 0 | 0 | — |
case-21 | pass→pass | 18,322 | 15,145 | -17% | 1 | 1 | 0% | 2,978 | 3,493 | +17% | 0 | 0 | — |
case-22 | pass→pass | 10,702 | 11,852 | +11% | 1 | 1 | 0% | 1,828 | 2,888 | +58% | 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 -20 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 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.