Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Map each candidate to the niches it covers, indicating strength of coverage for each assignment.
.claude/skills/yogsoth-ai-niche-mapping/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 1669% | 0% |
| case-11 | ✓→✓ | ＝ Same ✓ | 3% | 0% |
| case-12 | ✓→✓ | ＝ Same ✓ | 35% | 0% |
| case-13 | ✓→✓ | ＝ Same ✓ | 186% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | 55% | 0% |
Systematically assign candidates to the niches they cover, noting coverage strength and identifying which niches lack strong candidates.
Spawns a subagent that evaluates each candidate against each niche definition and produces a coverage map with gap identification.
Mapping requires evaluating each candidate against each niche criterion — a systematic cross-product analysis that benefits from methodical, uninterrupted execution.
Output must include a complete coverage map (every candidate assessed against every niche) and identification of gaps where no candidate provides strong coverage.
<!-- 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-01 | fail→fail | 27,241 | 36,669 | +35% | 1 | 1 | 0% | 3,777 | 5,852 | +55% | 0 | 0 | — |
case-02 | fail→fail | 6,502 | 64,445 | +891% | 1 | 1 | 0% | 1,082 | 7,810 | +622% | 0 | 0 | — |
case-03 | fail→fail | 11,620 | 31,397 | +170% | 1 | 1 | 0% | 2,104 | 5,044 | +140% | 0 | 0 | — |
case-04 | fail→fail | 11,739 | 35,201 | +200% | 1 | 1 | 0% | 1,081 | 6,561 | +507% | 0 | 0 | — |
case-05 | fail→fail | 14,739 | 37,867 | +157% | 1 | 1 | 0% | 1,788 | 6,769 | +279% | 0 | 0 | — |
case-06 | fail→fail | 11,496 | 35,700 | +211% | 1 | 1 | 0% | 1,819 | 5,253 | +189% | 0 | 0 | — |
case-07 | fail→fail | 15,684 | 45,827 | +192% | 1 | 1 | 0% | 1,728 | 7,572 | +338% | 0 | 0 | — |
case-08 | fail→fail | 11,616 | 29,114 | +151% | 1 | 1 | 0% | 1,064 | 936 | -12% | 0 | 0 | — |
case-09 | fail→fail | 21,229 | 36,649 | +73% | 1 | 1 | 0% | 2,624 | 3,276 | +25% | 0 | 0 | — |
case-10 | fail→pass | 2,177 | 32,072 | +1373% | 1 | 1 | 0% | 322 | 5,697 | +1669% | 0 | 0 | — |
case-11 | pass→pass | 24,360 | 22,546 | -7% | 1 | 1 | 0% | 2,952 | 3,035 | +3% | 0 | 0 | — |
case-12 | pass→pass | 21,992 | 24,013 | +9% | 1 | 1 | 0% | 2,493 | 3,360 | +35% | 0 | 0 | — |
case-13 | pass→pass | 7,752 | 32,218 | +316% | 1 | 1 | 0% | 1,638 | 4,684 | +186% | 0 | 0 | — |
case-14 | fail→fail | 20,938 | 29,227 | +40% | 1 | 1 | 0% | 2,878 | 6,216 | +116% | 0 | 0 | — |
case-15 | fail→fail | 13,376 | 31,045 | +132% | 1 | 1 | 0% | 2,256 | 4,753 | +111% | 0 | 0 | — |
case-16 | fail→fail | 21,231 | 33,625 | +58% | 1 | 1 | 0% | 3,453 | 5,248 | +52% | 0 | 0 | — |
case-17 | fail→fail | 17,444 | 28,579 | +64% | 1 | 1 | 0% | 2,096 | 4,348 | +107% | 0 | 0 | — |
case-18 | fail→fail | 15,122 | 37,311 | +147% | 1 | 1 | 0% | 2,556 | 5,952 | +133% | 0 | 0 | — |
case-19 | fail→fail | 12,087 | 33,998 | +181% | 1 | 1 | 0% | 1,164 | 5,763 | +395% | 0 | 0 | — |
case-20 | fail→fail | 15,422 | 40,129 | +160% | 1 | 1 | 0% | 2,575 | 6,867 | +167% | 0 | 0 | — |
case-21 | fail→fail | 23,138 | 32,622 | +41% | 1 | 1 | 0% | 2,885 | 4,812 | +67% | 0 | 0 | — |
case-22 | fail→fail | 23,950 | 30,917 | +29% | 1 | 1 | 0% | 3,221 | 4,587 | +42% | 0 | 0 | — |
case-23 | fail→fail | 17,662 | 31,409 | +78% | 1 | 1 | 0% | 2,156 | 5,093 | +136% | 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. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 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 +4 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.
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.