Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Identify research gaps via PICOS frameworks, concept matrices, evidence gap maps, keyword extraction, citation analysis, and topic modeling. Systematic discovery of what is missing in the literature.
.claude/skills/yogsoth-ai-deep-insight-gap-identification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 82% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 124% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 7% | 0% |
Discover where gaps exist in the research landscape.
User has a research area and needs to find what's missing — uncharted territory, under-explored intersections, or absent evidence.
| Base SOP | Target | ±10% Range | |----------|--------|------------| | web-search | 60 | 54–66 | | web-research | 10 | 9–11 | | paper-overview | 80 | 72–88 | | paper-search | 30 | 27–33 | | paper-research | 5 | 4–6 |
Print before every iteration:
<HARD-GATE>
| SOP | Done | Target | % |
|-----|------|--------|---|
| web-search | ? | 60 | ? |
| web-research | ? | 10 | ? |
| paper-overview | ? | 80 | ? |
| paper-search | ? | 30 | ? |
| paper-research | ? | 5 | ? |
Budget Gate: OPEN/CLOSED (>=80% required to exit)
</HARD-GATE>Import: web-search, web-research, paper-overview, paper-search, paper-research Subagent: gap-keyword-extraction, concept-matrix-construction, egm-construction
Start with broad web-search + paper-overview to map the landscape. Use gap-keyword-extraction to identify gap-indicating sentences. Build concept-matrix to visualize coverage. Construct EGM for structural view of where evidence is dense vs absent.
Gap Candidate List with evidence pointers — each candidate includes: gap description, evidence source, confidence level, and location in concept matrix.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | evidence-mapping | Systematic evidence map construction — search, classify, locate gaps, visualize. Combines concept-matrix-construction, gap-keyword-extraction, evidence-grading, and egm-construction SOPs. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | concept-matrix-construction | Build articles × concepts coverage matrix to visualize research landscape and identify empty cells as gap candidates. | | egm-construction | Build structured Evidence Gap Maps — define axes (intervention × outcome or method × domain), place gaps in cells, annotate with evidence density and quality. | | gap-keyword-extraction | Extract gap-indicating sentences and phrases from papers/reviews. Identifies linguistic markers of research gaps (e.g., "remains unclear", "has not been explored", "limited understanding"). |
<!-- 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 | 23,709 | 36,145 | +52% | 1 | 1 | 0% | 3,782 | 6,891 | +82% | 0 | 0 | — |
case-07 | pass→pass | 15,431 | 12,419 | -20% | 1 | 1 | 0% | 2,119 | 1,939 | -8% | 0 | 0 | — |
case-02 | fail→pass | 32,255 | 35,135 | +9% | 1 | 1 | 0% | 4,727 | 6,874 | +45% | 0 | 0 | — |
case-03 | fail→pass | 17,900 | 36,617 | +105% | 1 | 1 | 0% | 2,958 | 6,612 | +124% | 0 | 0 | — |
case-04 | fail→pass | 15,066 | 12,214 | -19% | 1 | 1 | 0% | 2,756 | 2,683 | -3% | 0 | 0 | — |
case-05 | fail→pass | 18,229 | 3,979 | -78% | 1 | 1 | 0% | 1,342 | 1,439 | +7% | 0 | 0 | — |
case-06 | fail→pass | 18,353 | 16,798 | -8% | 1 | 1 | 0% | 1,503 | 3,258 | +117% | 0 | 0 | — |
case-08 | pass→fail | 18,558 | 9,543 | -49% | 1 | 1 | 0% | 2,918 | 2,419 | -17% | 0 | 0 | — |
case-09 | pass→pass | 14,537 | 16,728 | +15% | 1 | 1 | 0% | 2,278 | 3,556 | +56% | 0 | 0 | — |
case-10 | pass→pass | 20,091 | 26,830 | +34% | 1 | 1 | 0% | 2,227 | 5,028 | +126% | 0 | 0 | — |
case-11 | fail→fail | 23,888 | 99,712 | +317% | 1 | 1 | 0% | 3,945 | 6,846 | +74% | 0 | 0 | — |
case-12 | pass→pass | 28,850 | 37,992 | +32% | 1 | 1 | 0% | 4,294 | 6,836 | +59% | 0 | 0 | — |
case-13 | pass→pass | 3,930 | 2,630 | -33% | 1 | 1 | 0% | 743 | 1,072 | +44% | 0 | 0 | — |
case-14 | pass→pass | 7,512 | 3,721 | -50% | 1 | 1 | 0% | 1,179 | 1,307 | +11% | 0 | 0 | — |
case-15 | pass→pass | 17,357 | 6,914 | -60% | 1 | 1 | 0% | 2,866 | 1,935 | -32% | 0 | 0 | — |
case-16 | fail→pass | 22,943 | 36,507 | +59% | 1 | 1 | 0% | 3,606 | 6,841 | +90% | 0 | 0 | — |
case-17 | fail→pass | 15,414 | 5,664 | -63% | 1 | 1 | 0% | 2,462 | 1,708 | -31% | 0 | 0 | — |
case-18 | pass→pass | 30,047 | 12,841 | -57% | 1 | 1 | 0% | 2,478 | 2,758 | +11% | 0 | 0 | — |
case-19 | fail→pass | 18,000 | 30,856 | +71% | 1 | 1 | 0% | 2,748 | 5,624 | +105% | 0 | 0 | — |
case-20 | pass→fail | 20,809 | 31,165 | +50% | 1 | 1 | 0% | 4,013 | 6,864 | +71% | 0 | 0 | — |
case-21 | pass→pass | 19,266 | 37,620 | +95% | 1 | 1 | 0% | 2,956 | 6,865 | +132% | 0 | 0 | — |
case-22 | pass→pass | 11,903 | 32,384 | +172% | 1 | 1 | 0% | 1,992 | 6,770 | +240% | 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 +32 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.