Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Strategy for defining ontology boundaries — identify seed concepts, classify topic size, establish scope constraints.
.claude/skills/yogsoth-ai-domain-scoping/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -8% | 0% |
Define the boundaries of the ontology. What's in scope, what's out, what are the seed concepts that anchor the domain.
Establish clear boundaries before extraction begins. An unbounded ontology grows without purpose. Seed concepts define the center of gravity; scope constraints define the edges.
| Metric | S | M | L | |--------|---|---|---| | Seed concepts identified | 5 | 10 | 20 | | Scope boundaries defined | 3 | 5 | 8 | | Source pages surveyed | 5 | 15 | 30 |
| Metric | Target | Current | Status |
|---------------------|--------|---------|--------|
| Seed concepts | X | 0 | ⬜ |
| Scope boundaries | X | 0 | ⬜ |
| Sources surveyed | X | 0 | ⬜ |<HARD-GATE> Cannot exit until 80% of budget met. Print state ledger before each iteration decision. </HARD-GATE>
After budget gate passes, ask:
Max 2 extra iterations if gaps found.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | concept-decomposition | Tactic for breaking compound concepts into atomic parts — split over-broad concepts, identify sub-components, create child pages. | | hierarchy-construction | Tactic for building is-a and part-of hierarchies — establish parent-child relationships, verify transitivity, detect cycles. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | seed-concept-search | SOP for finding seed concepts in existing vault and web sources to anchor ontology construction. |
<!-- 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-12 | fail→pass | 13,642 | 16,762 | +23% | 1 | 1 | 0% | 2,123 | 3,171 | +49% | 0 | 0 | — |
case-05 | fail→pass | 32,210 | 23,011 | -29% | 1 | 1 | 0% | 5,055 | 4,037 | -20% | 0 | 0 | — |
case-01 | fail→pass | 20,328 | 19,918 | -2% | 1 | 1 | 0% | 3,367 | 3,681 | +9% | 0 | 0 | — |
case-02 | fail→pass | 16,913 | 14,894 | -12% | 1 | 1 | 0% | 2,749 | 2,947 | +7% | 0 | 0 | — |
case-03 | fail→pass | 19,300 | 14,181 | -27% | 1 | 1 | 0% | 3,026 | 2,782 | -8% | 0 | 0 | — |
case-04 | fail→fail | 10,509 | 8,710 | -17% | 1 | 1 | 0% | 1,870 | 1,979 | +6% | 0 | 0 | — |
case-06 | fail→pass | 18,294 | 22,900 | +25% | 1 | 1 | 0% | 2,924 | 4,161 | +42% | 0 | 0 | — |
case-07 | pass→pass | 14,555 | 10,453 | -28% | 1 | 1 | 0% | 2,182 | 2,119 | -3% | 0 | 0 | — |
case-08 | pass→pass | 9,489 | 7,775 | -18% | 1 | 1 | 0% | 1,573 | 1,808 | +15% | 0 | 0 | — |
case-09 | pass→fail | 16,156 | 6,172 | -62% | 1 | 1 | 0% | 2,477 | 1,466 | -41% | 0 | 0 | — |
case-10 | pass→pass | 12,839 | 5,362 | -58% | 1 | 1 | 0% | 2,044 | 1,489 | -27% | 0 | 0 | — |
case-11 | fail→pass | 16,410 | 11,284 | -31% | 1 | 1 | 0% | 2,536 | 2,373 | -6% | 0 | 0 | — |
case-13 | pass→pass | 13,935 | 10,661 | -23% | 1 | 1 | 0% | 2,102 | 2,265 | +8% | 0 | 0 | — |
case-14 | fail→pass | 12,098 | 7,260 | -40% | 1 | 1 | 0% | 1,884 | 1,747 | -7% | 0 | 0 | — |
case-15 | pass→pass | 17,249 | 20,029 | +16% | 1 | 1 | 0% | 2,574 | 3,587 | +39% | 0 | 0 | — |
case-16 | pass→pass | 14,495 | 10,797 | -26% | 1 | 1 | 0% | 2,047 | 2,114 | +3% | 0 | 0 | — |
case-17 | fail→pass | 19,787 | 9,090 | -54% | 1 | 1 | 0% | 3,597 | 2,056 | -43% | 0 | 0 | — |
case-22 | pass→pass | 1,985 | 3,869 | +95% | 1 | 1 | 0% | 309 | 1,092 | +253% | 0 | 0 | — |
case-18 | fail→fail | 15,422 | 11,576 | -25% | 1 | 1 | 0% | 2,458 | 2,234 | -9% | 0 | 0 | — |
case-19 | fail→pass | 14,001 | 9,119 | -35% | 1 | 1 | 0% | 2,163 | 2,023 | -6% | 0 | 0 | — |
case-20 | pass→pass | 5,296 | 6,765 | +28% | 1 | 1 | 0% | 1,050 | 1,743 | +66% | 0 | 0 | — |
case-21 | pass→pass | 4,797 | 4,769 | -1% | 1 | 1 | 0% | 1,000 | 1,467 | +47% | 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 +41 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.