Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Drill from surface symptoms to root causes via 5 Whys, Ishikawa decomposition, and Current Reality Trees. Validates each causal link with literature evidence.
.claude/skills/yogsoth-ai-root-cause-drilling/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 133% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 140% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 186% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 197% | 0% |
Drill from surface phenomena to actionable root causes.
A gap or problem has been identified but the underlying cause is unclear — need to understand WHY it exists.
| Base SOP | Target | ±10% Range | |----------|--------|------------| | web-search | 30 | 27–33 | | web-research | 10 | 9–11 | | paper-overview | 40 | 36–44 | | paper-search | 25 | 22–28 | | paper-research | 10 | 9–11 |
<HARD-GATE>
| SOP | Done | Target | % |
|-----|------|--------|---|
| web-search | ? | 30 | ? |
| web-research | ? | 10 | ? |
| paper-overview | ? | 40 | ? |
| paper-search | ? | 25 | ? |
| paper-research | ? | 10 | ? |
Budget Gate: OPEN/CLOSED (>=80% required to exit)
</HARD-GATE>Import: web-search, web-research, paper-overview, paper-search, paper-research Subagent: five-whys-drilling, ishikawa-decomposition, current-reality-tree
Start with 5 Whys to drill from surface to root. Use Ishikawa for multi-factor decomposition. Build CRT for formal causal logic. Validate each causal link with literature evidence.
Root Cause Report — causal chain, Ishikawa diagram, CRT, validated root causes with evidence.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | causal-tree-building | Build logical causal trees from symptoms to root causes — list UDEs, connect causal chains, validate logic, locate root causes. Combines ishikawa-decomposition, current-reality-tree, and clr-validation SOPs. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | current-reality-tree | Build TOC Current Reality Trees — connect Undesirable Effects via sufficient-cause logic to identify 1-3 root causes. | | five-whys-drilling | Iterative "Why?" questioning (5+ levels) to drill from surface phenomenon to actionable root cause. Each level verified against evidence. | | ishikawa-decomposition | Decompose problems into 6M categories (Methodology, Data, Theory, Measurement, Researchers, Environment) via fishbone diagram analysis. |
<!-- 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-20 | pass→fail | 21,663 | 46,085 | +113% | 1 | 1 | 0% | 3,672 | 7,941 | +116% | 0 | 0 | — |
case-01 | fail→fail | 51,904 | 55,789 | +7% | 1 | 1 | 0% | 8,279 | 8,936 | +8% | 0 | 0 | — |
case-02 | fail→pass | 55,696 | 54,766 | -2% | 1 | 1 | 0% | 8,269 | 8,926 | +8% | 0 | 0 | — |
case-03 | fail→fail | 52,543 | 57,090 | +9% | 1 | 1 | 0% | 8,264 | 8,921 | +8% | 0 | 0 | — |
case-04 | fail→fail | 31,610 | 42,832 | +36% | 1 | 1 | 0% | 3,966 | 6,184 | +56% | 0 | 0 | — |
case-05 | fail→fail | 37,210 | 57,559 | +55% | 1 | 1 | 0% | 5,939 | 8,881 | +50% | 0 | 0 | — |
case-06 | fail→pass | 28,396 | 51,304 | +81% | 1 | 1 | 0% | 3,502 | 8,154 | +133% | 0 | 0 | — |
case-07 | fail→fail | 30,203 | 58,191 | +93% | 1 | 1 | 0% | 3,734 | 8,873 | +138% | 0 | 0 | — |
case-08 | fail→pass | 29,069 | 67,670 | +133% | 1 | 1 | 0% | 3,701 | 8,876 | +140% | 0 | 0 | — |
case-09 | fail→pass | 24,561 | 47,397 | +93% | 1 | 1 | 0% | 2,875 | 8,226 | +186% | 0 | 0 | — |
case-10 | fail→fail | 30,434 | 26,759 | -12% | 1 | 1 | 0% | 3,812 | 1,354 | -64% | 0 | 0 | — |
case-11 | fail→fail | 18,036 | 22,115 | +23% | 1 | 1 | 0% | 2,718 | 1,624 | -40% | 0 | 0 | — |
case-12 | fail→fail | 24,989 | 80,762 | +223% | 1 | 1 | 0% | 3,023 | 2,978 | -1% | 0 | 0 | — |
case-13 | fail→fail | 34,587 | 56,011 | +62% | 1 | 1 | 0% | 2,759 | 8,872 | +222% | 0 | 0 | — |
case-14 | fail→fail | 31,454 | 48,511 | +54% | 1 | 1 | 0% | 3,878 | 8,114 | +109% | 0 | 0 | — |
case-15 | fail→fail | 18,643 | 51,979 | +179% | 1 | 1 | 0% | 2,738 | 7,966 | +191% | 0 | 0 | — |
case-16 | fail→fail | 21,323 | 27,385 | +28% | 1 | 1 | 0% | 2,276 | 1,824 | -20% | 0 | 0 | — |
case-17 | fail→pass | 20,810 | 60,009 | +188% | 1 | 1 | 0% | 2,990 | 8,866 | +197% | 0 | 0 | — |
case-18 | fail→pass | 22,007 | 57,709 | +162% | 1 | 1 | 0% | 2,650 | 8,867 | +235% | 0 | 0 | — |
case-19 | fail→fail | 30,280 | 19,927 | -34% | 1 | 1 | 0% | 3,716 | 1,108 | -70% | 0 | 0 | — |
case-21 | pass→fail | 33,924 | 49,959 | +47% | 1 | 1 | 0% | 7,357 | 1,574 | -79% | 0 | 0 | — |
case-22 | pass→pass | 44,562 | 48,165 | +8% | 1 | 1 | 0% | 7,175 | 8,821 | +23% | 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 16 counted toward the lift figure. The other 6 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 +18 percentage points is the difference between those two pass rates over the 16 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.