Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Classify gaps using Miles 7-type taxonomy (theoretical, methodological, empirical, population, practical, knowledge void, evidence gap).
.claude/skills/yogsoth-ai-gap-typology-classification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✓→✓ | ＝ Same ✓ | 33% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | 37% | 0% |
| case-22 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
| case-06 | ✗→✗ | ＝ Same ✗ | -11% | 0% |
| case-04 | ✗→✗ | ＝ Same ✗ | 52% | 0% |
Classify research gaps by type using established taxonomies.
Subagent — spawned via subagent-spawning/spawn-agent.
One unit = one gap classification pass.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | spawn-agent | Spawn a customized CC subagent with full MCP tool access. Used by SOPs that declare execution: subagent. |
<!-- 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-06 | fail→fail | 13,926 | 12,333 | -11% | 1 | 1 | 0% | 2,331 | 2,085 | -11% | 0 | 0 | — |
case-04 | fail→fail | 9,818 | 19,392 | +98% | 1 | 1 | 0% | 1,620 | 2,462 | +52% | 0 | 0 | — |
case-05 | fail→fail | 18,752 | 17,996 | -4% | 1 | 1 | 0% | 2,885 | 3,099 | +7% | 0 | 0 | — |
case-01 | fail→fail | 10,349 | 17,777 | +72% | 1 | 1 | 0% | 1,689 | 1,528 | -10% | 0 | 0 | — |
case-02 | fail→fail | 17,637 | 28,511 | +62% | 1 | 1 | 0% | 2,775 | 4,585 | +65% | 0 | 0 | — |
case-03 | fail→fail | 8,013 | 11,414 | +42% | 1 | 1 | 0% | 1,291 | 1,919 | +49% | 0 | 0 | — |
case-07 | fail→fail | 6,232 | 9,205 | +48% | 1 | 1 | 0% | 980 | 1,554 | +59% | 0 | 0 | — |
case-08 | fail→fail | 15,065 | 14,072 | -7% | 1 | 1 | 0% | 2,366 | 2,238 | -5% | 0 | 0 | — |
case-09 | fail→fail | 19,876 | 31,315 | +58% | 1 | 1 | 0% | 3,179 | 4,994 | +57% | 0 | 0 | — |
case-10 | fail→fail | 7,829 | 14,768 | +89% | 1 | 1 | 0% | 1,182 | 2,519 | +113% | 0 | 0 | — |
case-11 | fail→fail | 7,084 | 9,402 | +33% | 1 | 1 | 0% | 1,042 | 1,743 | +67% | 0 | 0 | — |
case-12 | fail→fail | 22,110 | 17,342 | -22% | 1 | 1 | 0% | 3,396 | 2,736 | -19% | 0 | 0 | — |
case-13 | fail→fail | 9,143 | 20,634 | +126% | 1 | 1 | 0% | 1,427 | 3,530 | +147% | 0 | 0 | — |
case-14 | fail→fail | 15,489 | 20,517 | +32% | 1 | 1 | 0% | 2,553 | 3,247 | +27% | 0 | 0 | — |
case-15 | fail→fail | 17,104 | 17,919 | +5% | 1 | 1 | 0% | 2,766 | 2,815 | +2% | 0 | 0 | — |
case-16 | fail→fail | 19,641 | 15,547 | -21% | 1 | 1 | 0% | 2,994 | 2,638 | -12% | 0 | 0 | — |
case-17 | fail→fail | 12,282 | 16,650 | +36% | 1 | 1 | 0% | 1,866 | 2,724 | +46% | 0 | 0 | — |
case-18 | fail→fail | 11,504 | 17,496 | +52% | 1 | 1 | 0% | 1,699 | 2,885 | +70% | 0 | 0 | — |
case-19 | fail→fail | 12,093 | 15,303 | +27% | 1 | 1 | 0% | 1,931 | 2,462 | +27% | 0 | 0 | — |
case-20 | pass→pass | 17,736 | 21,940 | +24% | 1 | 1 | 0% | 2,627 | 3,489 | +33% | 0 | 0 | — |
case-21 | pass→pass | 7,920 | 9,900 | +25% | 1 | 1 | 0% | 1,367 | 1,867 | +37% | 0 | 0 | — |
case-22 | pass→pass | 11,998 | 14,900 | +24% | 1 | 1 | 0% | 2,499 | 3,077 | +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. The headline lift of 0 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.