Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Produce final structured audit report
.claude/skills/yogsoth-ai-benchmark-synthesis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 21% | 0% |
Synthesize all analysis results from a benchmark archaeology campaign into a final structured report with cross-cutting findings, prioritized recommendations, and actionable conclusions.
Final synthesis report suitable for publication or decision-making.
<!-- 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-22 | pass→pass | 17,540 | 14,608 | -17% | 1 | 1 | 0% | 3,347 | 3,132 | -6% | 0 | 0 | — |
case-01 | fail→pass | 19,409 | 15,781 | -19% | 1 | 1 | 0% | 2,865 | 2,405 | -16% | 0 | 0 | — |
case-02 | fail→pass | 22,295 | 35,895 | +61% | 1 | 1 | 0% | 3,340 | 5,488 | +64% | 0 | 0 | — |
case-03 | fail→fail | 17,754 | 15,099 | -15% | 1 | 1 | 0% | 2,610 | 2,346 | -10% | 0 | 0 | — |
case-04 | fail→pass | 19,513 | 23,494 | +20% | 1 | 1 | 0% | 3,069 | 3,716 | +21% | 0 | 0 | — |
case-05 | fail→fail | 11,608 | 7,909 | -32% | 1 | 1 | 0% | 1,906 | 1,410 | -26% | 0 | 0 | — |
case-06 | fail→fail | 24,529 | 31,679 | +29% | 1 | 1 | 0% | 3,663 | 5,535 | +51% | 0 | 0 | — |
case-07 | fail→pass | 6,909 | 6,837 | -1% | 1 | 1 | 0% | 993 | 1,174 | +18% | 0 | 0 | — |
case-08 | fail→fail | 17,145 | 23,869 | +39% | 1 | 1 | 0% | 2,321 | 3,525 | +52% | 0 | 0 | — |
case-09 | fail→fail | 18,216 | 15,062 | -17% | 1 | 1 | 0% | 2,448 | 2,311 | -6% | 0 | 0 | — |
case-10 | fail→fail | 9,567 | 9,601 | +0% | 1 | 1 | 0% | 1,505 | 1,714 | +14% | 0 | 0 | — |
case-11 | fail→fail | 10,817 | 17,059 | +58% | 1 | 1 | 0% | 1,363 | 2,601 | +91% | 0 | 0 | — |
case-12 | fail→pass | 44,730 | 28,368 | -37% | 1 | 1 | 0% | 3,784 | 4,583 | +21% | 0 | 0 | — |
case-13 | fail→pass | 5,075 | 13,359 | +163% | 1 | 1 | 0% | 707 | 2,081 | +194% | 0 | 0 | — |
case-14 | pass→pass | 8,487 | 12,194 | +44% | 1 | 1 | 0% | 1,196 | 1,974 | +65% | 0 | 0 | — |
case-15 | fail→fail | 20,014 | 18,537 | -7% | 1 | 1 | 0% | 2,983 | 2,856 | -4% | 0 | 0 | — |
case-16 | fail→fail | 19,443 | 25,033 | +29% | 1 | 1 | 0% | 2,718 | 3,828 | +41% | 0 | 0 | — |
case-17 | fail→fail | 15,687 | 16,415 | +5% | 1 | 1 | 0% | 2,145 | 2,506 | +17% | 0 | 0 | — |
case-18 | pass→pass | 16,859 | 15,302 | -9% | 1 | 1 | 0% | 2,363 | 2,463 | +4% | 0 | 0 | — |
case-19 | pass→pass | 22,184 | 21,692 | -2% | 1 | 1 | 0% | 3,022 | 3,109 | +3% | 0 | 0 | — |
case-20 | pass→pass | 14,810 | 25,119 | +70% | 1 | 1 | 0% | 2,332 | 3,999 | +71% | 0 | 0 | — |
case-21 | pass→pass | 32,443 | 30,311 | -7% | 1 | 1 | 0% | 6,176 | 6,401 | +4% | 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 +27 percentage points is the difference between those two pass rates over the 22 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.