Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Strategy for synthesizing argument positions — aggregate evidence, resolve contradictions, produce synthesis reports identifying which claims survive scrutiny.
.claude/skills/yogsoth-ai-argument-synthesis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 513% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 140% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 120% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 211% | 0% |
Synthesize coherent positions from the argument graph. Aggregates evidence across claims, resolves contradictions where possible, and produces synthesis reports that identify which claims survive scrutiny and which are undermined.
| Metric | Small | Medium | Large | |--------|-------|--------|-------| | Claims synthesized | 10 | 25 | 50 | | Contradictions resolved | 2 | 6 | 12 | | Synthesis reports | 1 | 3 | 5 |
<HARD-GATE> Print before every iteration:
| Metric | Target | Current | % | |--------|--------|---------|---| | Claims synthesized | — | — | — | | Contradictions resolved | — | — | — | | Synthesis reports | — | — | — | </HARD-GATE>
Cannot exit until 80% of budget targets met.
After budget gate passes, self-check:
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 | | --- | --- | | knowledge-structuring-claim-decomposition | Tactic for decomposing compound claims into atomic propositions — identify logical structure, separate conjunctions, extract implicit assumptions. | | strength-assessment | Tactic for assessing argument strength — evaluate evidence quality, count independent sources, check for defeaters, assign calibrated confidence scores. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | argument-visualization | SOP for generating argument structure visualization — query graph for argument chains, format as mermaid diagram or indented tree, write to vault. | | synthesis-report | SOP for producing argument synthesis reports — aggregate evidence, resolve contradictions, identify surviving claims, write structured summary. |
<!-- 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 | 7,554 | 40,346 | +434% | 1 | 1 | 0% | 1,097 | 6,729 | +513% | 0 | 0 | — |
case-02 | fail→pass | 20,286 | 40,592 | +100% | 1 | 1 | 0% | 2,801 | 6,726 | +140% | 0 | 0 | — |
case-03 | fail→pass | 33,156 | 20,982 | -37% | 1 | 1 | 0% | 4,842 | 3,664 | -24% | 0 | 0 | — |
case-04 | fail→pass | 13,306 | 23,559 | +77% | 1 | 1 | 0% | 1,964 | 4,317 | +120% | 0 | 0 | — |
case-05 | fail→pass | 14,269 | 40,350 | +183% | 1 | 1 | 0% | 2,151 | 6,699 | +211% | 0 | 0 | — |
case-06 | fail→fail | 29,403 | 37,875 | +29% | 1 | 1 | 0% | 4,495 | 6,693 | +49% | 0 | 0 | — |
case-07 | fail→pass | 3,748 | 5,045 | +35% | 1 | 1 | 0% | 648 | 1,466 | +126% | 0 | 0 | — |
case-08 | fail→pass | 14,213 | 4,297 | -70% | 1 | 1 | 0% | 1,963 | 1,274 | -35% | 0 | 0 | — |
case-09 | fail→pass | 11,706 | 4,053 | -65% | 1 | 1 | 0% | 1,606 | 1,145 | -29% | 0 | 0 | — |
case-10 | fail→pass | 7,561 | 1,767 | -77% | 1 | 1 | 0% | 999 | 783 | -22% | 0 | 0 | — |
case-11 | pass→pass | 6,485 | 29,909 | +361% | 1 | 1 | 0% | 1,121 | 5,554 | +395% | 0 | 0 | — |
case-12 | pass→fail | 17,205 | 8,272 | -52% | 1 | 1 | 0% | 2,477 | 1,953 | -21% | 0 | 0 | — |
case-13 | pass→pass | 17,932 | 38,278 | +113% | 1 | 1 | 0% | 2,918 | 6,703 | +130% | 0 | 0 | — |
case-14 | pass→pass | 20,176 | 38,902 | +93% | 1 | 1 | 0% | 3,076 | 6,696 | +118% | 0 | 0 | — |
case-15 | fail→pass | 10,904 | 9,893 | -9% | 1 | 1 | 0% | 1,574 | 2,169 | +38% | 0 | 0 | — |
case-20 | fail→pass | 1,669 | 30,886 | +1751% | 1 | 1 | 0% | 202 | 5,834 | +2788% | 0 | 0 | — |
case-16 | fail→pass | 16,438 | 39,391 | +140% | 1 | 1 | 0% | 2,479 | 6,690 | +170% | 0 | 0 | — |
case-17 | pass→fail | 11,467 | 6,109 | -47% | 1 | 1 | 0% | 1,494 | 1,565 | +5% | 0 | 0 | — |
case-18 | fail→pass | 13,550 | 2,007 | -85% | 1 | 1 | 0% | 1,944 | 794 | -59% | 0 | 0 | — |
case-19 | pass→pass | 13,586 | 11,056 | -19% | 1 | 1 | 0% | 1,888 | 2,327 | +23% | 0 | 0 | — |
case-21 | fail→fail | 2,008 | 24,757 | +1133% | 1 | 1 | 0% | 314 | 4,651 | +1381% | 0 | 0 | — |
case-22 | pass→fail | 5,492 | 15,380 | +180% | 1 | 1 | 0% | 823 | 3,033 | +269% | 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 +45 percentage points is the difference between those two pass rates over the 22 comparable cases. 4 cases got worse with the skill loaded, and they are 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.