Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Strategy for mining concepts from sources — systematic extraction, page creation, alias resolution.
.claude/skills/yogsoth-ai-concept-extraction/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 237% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 30% | 0% |
Mine source pages for concepts. Create concept pages, resolve aliases, ensure every concept is properly named and connected.
Extract concepts systematically from sources. Prefer precision over recall — it's better to have 20 well-defined concepts than 50 vague ones. Resolve naming conflicts immediately.
| Metric | S | M | L | |--------|---|---|---| | Sources processed | 8 | 20 | 40 | | Concepts extracted | 12 | 30 | 60 | | Aliases resolved | 3 | 10 | 20 |
| Metric | Target | Current | Status |
|---------------------|--------|---------|--------|
| Sources processed | X | 0 | ⬜ |
| Concepts extracted | X | 0 | ⬜ |
| Aliases resolved | X | 0 | ⬜ |<HARD-GATE> Cannot exit until 80% of budget met. Print state ledger before each iteration decision. </HARD-GATE>
<!-- 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. | | knowledge-structuring-consistency-checking | Tactic for verifying ontology consistency — detect contradictions, cycles, orphans, and type violations. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | source-gathering | SOP for gathering source material relevant to the ontology domain — fetch papers and web pages into source pages. |
<!-- 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 | 6,423 | 16,532 | +157% | 1 | 1 | 0% | 1,065 | 3,593 | +237% | 0 | 0 | — |
case-02 | fail→fail | 33,737 | 9,785 | -71% | 1 | 1 | 0% | 6,191 | 2,117 | -66% | 0 | 0 | — |
case-03 | fail→pass | 18,470 | 13,900 | -25% | 1 | 1 | 0% | 3,423 | 3,291 | -4% | 0 | 0 | — |
case-04 | pass→pass | 13,999 | 9,845 | -30% | 1 | 1 | 0% | 2,166 | 2,174 | +0% | 0 | 0 | — |
case-05 | fail→pass | 18,950 | 2,680 | -86% | 1 | 1 | 0% | 972 | 871 | -10% | 0 | 0 | — |
case-06 | fail→pass | 14,797 | 2,872 | -81% | 1 | 1 | 0% | 2,490 | 985 | -60% | 0 | 0 | — |
case-07 | fail→pass | 11,092 | 2,680 | -76% | 1 | 1 | 0% | 743 | 965 | +30% | 0 | 0 | — |
case-08 | pass→pass | 11,513 | 2,608 | -77% | 1 | 1 | 0% | 1,805 | 815 | -55% | 0 | 0 | — |
case-09 | pass→pass | 8,069 | 2,389 | -70% | 1 | 1 | 0% | 1,290 | 876 | -32% | 0 | 0 | — |
case-10 | pass→pass | 7,163 | 1,755 | -75% | 1 | 1 | 0% | 1,068 | 689 | -35% | 0 | 0 | — |
case-21 | pass→pass | 13,940 | 18,206 | +31% | 1 | 1 | 0% | 2,493 | 3,854 | +55% | 0 | 0 | — |
case-11 | fail→pass | 9,681 | 4,364 | -55% | 1 | 1 | 0% | 923 | 1,277 | +38% | 0 | 0 | — |
case-12 | pass→pass | 7,298 | 3,395 | -53% | 1 | 1 | 0% | 1,295 | 1,053 | -19% | 0 | 0 | — |
case-13 | fail→pass | 9,984 | 3,022 | -70% | 1 | 1 | 0% | 1,515 | 975 | -36% | 0 | 0 | — |
case-14 | fail→pass | 11,004 | 3,137 | -71% | 1 | 1 | 0% | 1,618 | 927 | -43% | 0 | 0 | — |
case-15 | pass→pass | 14,887 | 14,005 | -6% | 1 | 1 | 0% | 2,279 | 2,968 | +30% | 0 | 0 | — |
case-16 | pass→pass | 13,791 | 3,366 | -76% | 1 | 1 | 0% | 2,042 | 914 | -55% | 0 | 0 | — |
case-17 | fail→fail | 9,655 | 2,819 | -71% | 1 | 1 | 0% | 1,479 | 976 | -34% | 0 | 0 | — |
case-18 | pass→pass | 11,489 | 7,351 | -36% | 1 | 1 | 0% | 1,649 | 1,547 | -6% | 0 | 0 | — |
case-19 | fail→fail | 8,862 | 4,611 | -48% | 1 | 1 | 0% | 1,258 | 1,205 | -4% | 0 | 0 | — |
case-20 | pass→fail | 18,620 | 27,188 | +46% | 1 | 1 | 0% | 2,748 | 5,058 | +84% | 0 | 0 | — |
case-22 | pass→pass | 6,152 | 5,547 | -10% | 1 | 1 | 0% | 907 | 1,337 | +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, and 20 counted toward the lift figure. The other 2 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 +32 percentage points is the difference between those two pass rates over the 20 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.