Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Classify identified gaps using Miles 7-type taxonomy and AHRQ 4-reason framework. Determines gap type (theoretical, methodological, empirical, etc.) and root cause of gap existence.
.claude/skills/yogsoth-ai-gap-classification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 117% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 48% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 72% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 169% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 137% | 0% |
Classify identified gaps by type and root cause.
Gap candidates have been identified and need systematic categorization to inform prioritization and resolution strategy.
| Base SOP | Target | ±10% Range | |----------|--------|------------| | web-search | 20 | 18–22 | | web-research | 5 | 4–6 | | paper-overview | 30 | 27–33 | | paper-search | 20 | 18–22 | | paper-research | 10 | 9–11 |
Print before every iteration:
<HARD-GATE>
| SOP | Done | Target | % |
|-----|------|--------|---|
| web-search | ? | 20 | ? |
| web-research | ? | 5 | ? |
| paper-overview | ? | 30 | ? |
| paper-search | ? | 20 | ? |
| 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: gap-typology-classification, ahrq-reason-classification
For each gap candidate, apply Miles 7-type taxonomy (theoretical, methodological, empirical, population, practical, knowledge void, evidence gap), then AHRQ 4-reason classification for root cause of gap existence (insufficient info, biased info, inconsistent info, not yet integrated).
Classified Gap Table — each gap with: type label, AHRQ reason, confidence, supporting evidence.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | ahrq-reason-classification | Classify gap root causes using AHRQ 4-reason framework (insufficient info, biased info, inconsistent info, not yet integrated). | | gap-typology-classification | Classify gaps using Miles 7-type taxonomy (theoretical, methodological, empirical, population, practical, knowledge void, evidence gap). |
<!-- 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-04 | pass→fail | 15,745 | 36,777 | +134% | 1 | 1 | 0% | 2,411 | 6,717 | +179% | 0 | 0 | — |
case-01 | fail→pass | 17,019 | 29,945 | +76% | 1 | 1 | 0% | 2,654 | 5,748 | +117% | 0 | 0 | — |
case-02 | fail→fail | 15,318 | 6,686 | -56% | 1 | 1 | 0% | 2,440 | 1,831 | -25% | 0 | 0 | — |
case-03 | fail→pass | 24,227 | 29,063 | +20% | 1 | 1 | 0% | 3,768 | 5,588 | +48% | 0 | 0 | — |
case-05 | pass→pass | 17,639 | 27,199 | +54% | 1 | 1 | 0% | 3,178 | 5,922 | +86% | 0 | 0 | — |
case-06 | pass→fail | 17,486 | 31,917 | +83% | 1 | 1 | 0% | 3,180 | 6,716 | +111% | 0 | 0 | — |
case-07 | fail→pass | 13,859 | 19,362 | +40% | 1 | 1 | 0% | 2,268 | 3,896 | +72% | 0 | 0 | — |
case-08 | fail→pass | 9,689 | 23,212 | +140% | 1 | 1 | 0% | 1,653 | 4,441 | +169% | 0 | 0 | — |
case-09 | fail→pass | 8,430 | 17,429 | +107% | 1 | 1 | 0% | 1,522 | 3,606 | +137% | 0 | 0 | — |
case-10 | fail→pass | 10,500 | 15,075 | +44% | 1 | 1 | 0% | 1,661 | 3,061 | +84% | 0 | 0 | — |
case-11 | fail→pass | 11,558 | 30,564 | +164% | 1 | 1 | 0% | 1,926 | 6,055 | +214% | 0 | 0 | — |
case-12 | fail→pass | 10,620 | 13,164 | +24% | 1 | 1 | 0% | 1,612 | 2,878 | +79% | 0 | 0 | — |
case-13 | fail→pass | 10,404 | 16,986 | +63% | 1 | 1 | 0% | 1,709 | 3,644 | +113% | 0 | 0 | — |
case-14 | fail→pass | 15,856 | 21,735 | +37% | 1 | 1 | 0% | 2,595 | 4,359 | +68% | 0 | 0 | — |
case-15 | fail→pass | 10,479 | 17,414 | +66% | 1 | 1 | 0% | 1,578 | 3,621 | +129% | 0 | 0 | — |
case-16 | fail→fail | 12,253 | 6,725 | -45% | 1 | 1 | 0% | 2,100 | 1,630 | -22% | 0 | 0 | — |
case-17 | fail→pass | 12,358 | 15,553 | +26% | 1 | 1 | 0% | 1,954 | 3,204 | +64% | 0 | 0 | — |
case-18 | fail→pass | 12,293 | 18,219 | +48% | 1 | 1 | 0% | 1,911 | 3,808 | +99% | 0 | 0 | — |
case-19 | fail→pass | 10,558 | 12,393 | +17% | 1 | 1 | 0% | 1,867 | 2,688 | +44% | 0 | 0 | — |
case-20 | fail→fail | 13,142 | 8,703 | -34% | 1 | 1 | 0% | 2,103 | 1,965 | -7% | 0 | 0 | — |
case-21 | fail→pass | 12,046 | 19,205 | +59% | 1 | 1 | 0% | 1,773 | 3,725 | +110% | 0 | 0 | — |
case-22 | fail→pass | 13,557 | 26,278 | +94% | 1 | 1 | 0% | 2,105 | 4,741 | +125% | 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 +64 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 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.