Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Paper AI summary reading — deeper understanding of specific methodology papers without full-text commitment.
.claude/skills/yogsoth-ai-convergence-paper-search/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -64% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 17% | 0% |
Import SOP. Execution delegated to literature-engine/paper-search.
See source repo for full SKILL.md and prompt.md.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | literature-search | Medium-depth literature search — read AI-summarized reports for every paper analyzed |
<!-- 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 | 18,026 | 31,456 | +75% | 1 | 1 | 0% | 2,906 | 5,469 | +88% | 0 | 0 | — |
case-02 | fail→fail | 24,792 | 39,182 | +58% | 1 | 1 | 0% | 4,281 | 6,280 | +47% | 0 | 0 | — |
case-03 | fail→fail | 20,813 | 34,404 | +65% | 1 | 1 | 0% | 3,472 | 5,572 | +60% | 0 | 0 | — |
case-04 | fail→fail | 35,071 | 1,999 | -94% | 1 | 1 | 0% | 1,112 | 413 | -63% | 0 | 0 | — |
case-05 | fail→fail | 17,521 | 43,840 | +150% | 1 | 1 | 0% | 2,865 | 3,890 | +36% | 0 | 0 | — |
case-06 | fail→pass | 35,984 | 7,326 | -80% | 1 | 1 | 0% | 3,450 | 1,233 | -64% | 0 | 0 | — |
case-07 | fail→pass | 17,519 | 13,878 | -21% | 1 | 1 | 0% | 2,533 | 2,373 | -6% | 0 | 0 | — |
case-08 | fail→pass | 14,997 | 35,247 | +135% | 1 | 1 | 0% | 2,654 | 3,207 | +21% | 0 | 0 | — |
case-09 | fail→fail | 15,685 | 30,641 | +95% | 1 | 1 | 0% | 3,090 | 5,016 | +62% | 0 | 0 | — |
case-10 | fail→pass | 9,022 | 10,120 | +12% | 1 | 1 | 0% | 1,796 | 2,239 | +25% | 0 | 0 | — |
case-11 | fail→fail | 16,699 | 22,470 | +35% | 1 | 1 | 0% | 2,655 | 3,937 | +48% | 0 | 0 | — |
case-12 | fail→fail | 14,225 | 11,756 | -17% | 1 | 1 | 0% | 2,763 | 1,878 | -32% | 0 | 0 | — |
case-13 | fail→pass | 20,033 | 22,696 | +13% | 1 | 1 | 0% | 3,229 | 3,785 | +17% | 0 | 0 | — |
case-14 | fail→fail | 19,652 | 20,512 | +4% | 1 | 1 | 0% | 3,063 | 4,233 | +38% | 0 | 0 | — |
case-15 | fail→pass | 15,973 | 15,109 | -5% | 1 | 1 | 0% | 2,952 | 3,043 | +3% | 0 | 0 | — |
case-16 | fail→pass | 18,590 | 25,009 | +35% | 1 | 1 | 0% | 3,308 | 4,905 | +48% | 0 | 0 | — |
case-17 | fail→pass | 8,583 | 15,375 | +79% | 1 | 1 | 0% | 1,448 | 2,439 | +68% | 0 | 0 | — |
case-18 | fail→pass | 13,745 | 16,206 | +18% | 1 | 1 | 0% | 2,201 | 3,050 | +39% | 0 | 0 | — |
case-19 | fail→pass | 16,542 | 18,523 | +12% | 1 | 1 | 0% | 2,654 | 2,975 | +12% | 0 | 0 | — |
case-20 | pass→pass | 3,701 | 4,253 | +15% | 1 | 1 | 0% | 486 | 754 | +55% | 0 | 0 | — |
case-21 | pass→pass | 5,683 | 6,456 | +14% | 1 | 1 | 0% | 1,294 | 1,481 | +14% | 0 | 0 | — |
case-22 | pass→pass | 33,388 | 17,944 | -46% | 1 | 1 | 0% | 4,140 | 4,602 | +11% | 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 21 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 +45 percentage points is the difference between those two pass rates over the 21 comparable cases.
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.