Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Execute multi-objective optimization on candidates to produce a Pareto front of non-dominated solutions.
.claude/skills/yogsoth-ai-optimization-run/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✓→✗ | ▼ Worse | 214% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 103% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 18% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | 383% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | -29% | 0% |
Execute multi-objective optimization over the candidate set given defined objectives and constraints, producing a Pareto front of non-dominated solutions.
Spawns a subagent that applies optimization logic to enumerate, evaluate, and filter candidate portfolios, returning the non-dominated set.
Optimization requires systematic enumeration or heuristic search across the combinatorial space of possible portfolios. This computational work is self-contained and produces a well-defined output structure.
Output must contain at least 5 non-dominated solutions on the Pareto front. If the candidate set is too small or constraints too tight, report the maximum achievable frontier size with explanation.
<!-- 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-01 | fail→fail | 14,087 | 51,778 | +268% | 1 | 1 | 0% | 1,492 | 7,203 | +383% | 0 | 0 | — |
case-02 | fail→fail | 6,542 | 12,964 | +98% | 1 | 1 | 0% | 1,093 | 778 | -29% | 0 | 0 | — |
case-03 | pass→pass | 22,794 | 82,945 | +264% | 1 | 1 | 0% | 4,152 | 8,449 | +103% | 0 | 0 | — |
case-04 | pass→pass | 18,488 | 17,657 | -4% | 1 | 1 | 0% | 3,259 | 3,861 | +18% | 0 | 0 | — |
case-05 | pass→fail | 18,243 | 42,920 | +135% | 1 | 1 | 0% | 2,689 | 8,453 | +214% | 0 | 0 | — |
case-06 | fail→fail | 6,321 | 28,637 | +353% | 1 | 1 | 0% | 1,034 | 4,714 | +356% | 0 | 0 | — |
case-07 | fail→fail | 9,363 | 29,255 | +212% | 1 | 1 | 0% | 670 | 858 | +28% | 0 | 0 | — |
case-08 | fail→fail | 26,797 | 48,525 | +81% | 1 | 1 | 0% | 3,156 | 8,446 | +168% | 0 | 0 | — |
case-09 | fail→fail | 20,157 | 131,652 | +553% | 1 | 1 | 0% | 3,745 | 8,450 | +126% | 0 | 0 | — |
case-10 | fail→fail | 18,921 | 72,052 | +281% | 1 | 1 | 0% | 3,193 | 8,436 | +164% | 0 | 0 | — |
case-11 | fail→fail | 5,609 | 20,116 | +259% | 1 | 1 | 0% | 1,079 | 4,458 | +313% | 0 | 0 | — |
case-12 | fail→fail | 17,376 | 39,039 | +125% | 1 | 1 | 0% | 2,867 | 8,434 | +194% | 0 | 0 | — |
case-13 | fail→fail | 20,129 | 37,977 | +89% | 1 | 1 | 0% | 3,367 | 8,123 | +141% | 0 | 0 | — |
case-14 | fail→fail | 21,187 | 49,815 | +135% | 1 | 1 | 0% | 3,386 | 8,435 | +149% | 0 | 0 | — |
case-15 | fail→fail | 22,461 | 38,967 | +73% | 1 | 1 | 0% | 3,790 | 8,434 | +123% | 0 | 0 | — |
case-16 | fail→fail | 19,655 | 49,066 | +150% | 1 | 1 | 0% | 3,102 | 8,435 | +172% | 0 | 0 | — |
case-17 | fail→fail | 24,001 | 120,494 | +402% | 1 | 1 | 0% | 4,054 | 8,440 | +108% | 0 | 0 | — |
case-18 | fail→fail | 15,927 | 85,496 | +437% | 1 | 1 | 0% | 2,740 | 8,430 | +208% | 0 | 0 | — |
case-19 | fail→fail | 16,775 | 8,245 | -51% | 1 | 1 | 0% | 2,606 | 608 | -77% | 0 | 0 | — |
case-20 | fail→fail | 31,817 | 38,295 | +20% | 1 | 1 | 0% | 3,217 | 8,436 | +162% | 0 | 0 | — |
case-21 | fail→fail | 9,775 | 48,097 | +392% | 1 | 1 | 0% | 1,459 | 3,828 | +162% | 0 | 0 | — |
case-22 | fail→fail | 35,528 | 36,762 | +3% | 1 | 1 | 0% | 3,761 | 8,434 | +124% | 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 19 counted toward the lift figure. The other 3 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 -5 percentage points is the difference between those two pass rates over the 19 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.