Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Classic Zwicky box: parameter identification → value enumeration → matrix construction
.claude/skills/yogsoth-ai-zwicky-box-construction/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 122% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 8% | 0% |
Classic Zwicky morphological box: identify parameters, enumerate values, construct matrix, generate combinations.
| Resource | Target | Current | % | |----------|--------|---------|---| | web-search | 25 | 0 | 0% | | web-research | 8 | 0 | 0% | | paper-overview | 25 | 0 | 0% | | paper-search | 15 | 0 | 0% | | paper-research | 5 | 0 | 0% |
Cannot exit strategy until ≥80% of each budget line is consumed OR yield targets are met with justification for remaining budget.
| Tactic | Role | |--------|------| | combination-mapping | Enumerate parameter × value combinations | | consistency-checking | Filter inconsistent combinations |
| SOP | Role | |-----|------| | value-enumeration | Enumerate 3-5 values per parameter | | matrix-construction | Build the morphological matrix | | path-generation | Generate combination paths | | morphological-synthesis | Synthesize final output |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | matrix-construction | Build n-dimensional morphological matrix | | morphological-synthesis | Synthesize all morphological exploration outputs | | path-generation | Generate combination paths through consistent space | | value-enumeration | Enumerate 3-5 values per parameter including extremes |
<!-- 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-03 | pass→pass | 15,977 | 17,710 | +11% | 1 | 1 | 0% | 1,736 | 2,453 | +41% | 0 | 0 | — |
case-01 | pass→pass | 21,908 | 35,223 | +61% | 1 | 1 | 0% | 2,852 | 6,308 | +121% | 0 | 0 | — |
case-02 | fail→pass | 21,655 | 12,533 | -42% | 1 | 1 | 0% | 2,528 | 1,728 | -32% | 0 | 0 | — |
case-04 | pass→pass | 16,541 | 34,537 | +109% | 1 | 1 | 0% | 1,782 | 5,874 | +230% | 0 | 0 | — |
case-05 | pass→pass | 17,697 | 20,444 | +16% | 1 | 1 | 0% | 2,059 | 3,376 | +64% | 0 | 0 | — |
case-06 | pass→pass | 13,785 | 15,307 | +11% | 1 | 1 | 0% | 1,396 | 2,163 | +55% | 0 | 0 | — |
case-07 | fail→fail | 10,649 | 14,173 | +33% | 1 | 1 | 0% | 923 | 2,065 | +124% | 0 | 0 | — |
case-08 | pass→pass | 19,525 | 22,424 | +15% | 1 | 1 | 0% | 2,345 | 2,832 | +21% | 0 | 0 | — |
case-15 | pass→pass | 22,172 | 29,382 | +33% | 1 | 1 | 0% | 3,047 | 4,953 | +63% | 0 | 0 | — |
case-09 | fail→pass | 13,196 | 9,016 | -32% | 1 | 1 | 0% | 1,320 | 1,148 | -13% | 0 | 0 | — |
case-10 | pass→pass | 17,697 | 11,474 | -35% | 1 | 1 | 0% | 1,912 | 1,459 | -24% | 0 | 0 | — |
case-11 | fail→pass | 17,838 | 24,427 | +37% | 1 | 1 | 0% | 2,132 | 3,739 | +75% | 0 | 0 | — |
case-12 | pass→pass | 22,252 | 36,661 | +65% | 1 | 1 | 0% | 2,506 | 5,777 | +131% | 0 | 0 | — |
case-13 | fail→pass | 15,501 | 22,354 | +44% | 1 | 1 | 0% | 1,671 | 3,703 | +122% | 0 | 0 | — |
case-14 | fail→pass | 17,741 | 13,546 | -24% | 1 | 1 | 0% | 1,909 | 2,062 | +8% | 0 | 0 | — |
case-16 | pass→pass | 19,984 | 17,302 | -13% | 1 | 1 | 0% | 2,272 | 2,443 | +8% | 0 | 0 | — |
case-17 | pass→pass | 21,815 | 26,477 | +21% | 1 | 1 | 0% | 2,696 | 4,226 | +57% | 0 | 0 | — |
case-18 | fail→pass | 20,357 | 20,816 | +2% | 1 | 1 | 0% | 2,256 | 3,065 | +36% | 0 | 0 | — |
case-19 | pass→pass | 10,129 | 10,172 | +0% | 1 | 1 | 0% | 986 | 1,425 | +45% | 0 | 0 | — |
case-20 | pass→fail | 20,839 | 41,065 | +97% | 1 | 1 | 0% | 2,510 | 6,110 | +143% | 0 | 0 | — |
case-21 | pass→fail | 25,456 | 45,467 | +79% | 1 | 1 | 0% | 3,222 | 7,472 | +132% | 0 | 0 | — |
case-22 | pass→fail | 19,213 | 35,093 | +83% | 1 | 1 | 0% | 2,344 | 5,614 | +140% | 0 | 0 | — |
case-23 | pass→fail | 20,540 | 40,073 | +95% | 1 | 1 | 0% | 2,583 | 6,542 | +153% | 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. 23 cases were attempted. The headline lift of +9 percentage points is the difference between those two pass rates over the 23 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.