Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Strategy for identifying fundamental dimensions of variation in a design space.
.claude/skills/yogsoth-ai-dimension-discovery/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 83% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 7% | 0% |
Identify the fundamental axes of variation that define the design space. Each dimension should be independent and meaningful.
Look for the axes that practitioners implicitly use when comparing approaches. "Fast vs accurate", "local vs global", "supervised vs unsupervised" — these are dimensions. Find the ones specific to your domain.
| Metric | S | M | L | |--------|---|---|---| | Candidate dimensions | 5 | 10 | 15 | | Validated dimensions | 3 | 6 | 10 | | Sources analyzed | 8 | 20 | 40 |
| Metric | Target | Current | Status |
|---------------------|--------|---------|--------|
| Candidate dimensions| X | 0 | ⬜ |
| Validated dimensions| X | 0 | ⬜ |
| Sources analyzed | X | 0 | ⬜ |<HARD-GATE> Cannot exit until 80% of budget met. Print state ledger before each iteration decision. </HARD-GATE>
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | axis-extraction | Tactic for systematically extracting axes of variation from literature — identify how practitioners compare approaches. | | matrix-generation | Tactic for generating and populating combination matrices — cross dimensions to enumerate the design space. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | dimension-page-creation | SOP for creating a dimension page — documents an axis of variation with its values and semantics. |
<!-- 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 | 21,650 | 37,498 | +73% | 1 | 1 | 0% | 3,628 | 6,626 | +83% | 0 | 0 | — |
case-02 | fail→pass | 25,798 | 21,164 | -18% | 1 | 1 | 0% | 4,122 | 3,690 | -10% | 0 | 0 | — |
case-03 | fail→pass | 20,068 | 25,003 | +25% | 1 | 1 | 0% | 3,063 | 4,558 | +49% | 0 | 0 | — |
case-04 | pass→fail | 17,525 | 12,052 | -31% | 1 | 1 | 0% | 2,713 | 2,301 | -15% | 0 | 0 | — |
case-05 | pass→fail | 26,576 | 31,109 | +17% | 1 | 1 | 0% | 4,487 | 5,604 | +25% | 0 | 0 | — |
case-06 | pass→fail | 13,793 | 24,428 | +77% | 1 | 1 | 0% | 1,982 | 4,165 | +110% | 0 | 0 | — |
case-07 | pass→pass | 13,368 | 19,839 | +48% | 1 | 1 | 0% | 2,241 | 3,466 | +55% | 0 | 0 | — |
case-08 | pass→pass | 18,602 | 23,974 | +29% | 1 | 1 | 0% | 2,910 | 4,257 | +46% | 0 | 0 | — |
case-09 | fail→pass | 31,711 | 35,360 | +12% | 1 | 1 | 0% | 4,779 | 6,082 | +27% | 0 | 0 | — |
case-10 | fail→pass | 40,564 | 37,647 | -7% | 1 | 1 | 0% | 6,165 | 6,608 | +7% | 0 | 0 | — |
case-11 | fail→pass | 18,437 | 10,984 | -40% | 1 | 1 | 0% | 3,086 | 2,110 | -32% | 0 | 0 | — |
case-12 | fail→pass | 19,144 | 39,019 | +104% | 1 | 1 | 0% | 3,087 | 6,609 | +114% | 0 | 0 | — |
case-13 | fail→pass | 21,624 | 19,722 | -9% | 1 | 1 | 0% | 3,405 | 3,811 | +12% | 0 | 0 | — |
case-14 | pass→pass | 23,355 | 29,757 | +27% | 1 | 1 | 0% | 3,670 | 5,245 | +43% | 0 | 0 | — |
case-15 | fail→pass | 19,464 | 37,750 | +94% | 1 | 1 | 0% | 3,043 | 6,610 | +117% | 0 | 0 | — |
case-16 | fail→pass | 33,782 | 37,338 | +11% | 1 | 1 | 0% | 6,169 | 6,380 | +3% | 0 | 0 | — |
case-17 | fail→pass | 34,758 | 37,939 | +9% | 1 | 1 | 0% | 5,630 | 6,603 | +17% | 0 | 0 | — |
case-18 | fail→pass | 21,348 | 30,727 | +44% | 1 | 1 | 0% | 3,303 | 5,470 | +66% | 0 | 0 | — |
case-19 | pass→pass | 29,339 | 40,162 | +37% | 1 | 1 | 0% | 4,628 | 6,603 | +43% | 0 | 0 | — |
case-20 | fail→pass | 21,420 | 29,518 | +38% | 1 | 1 | 0% | 3,255 | 5,270 | +62% | 0 | 0 | — |
case-21 | pass→pass | 24,111 | 27,414 | +14% | 1 | 1 | 0% | 3,752 | 4,799 | +28% | 0 | 0 | — |
case-22 | fail→pass | 24,053 | 27,915 | +16% | 1 | 1 | 0% | 3,657 | 4,937 | +35% | 0 | 0 | — |
case-23 | fail→pass | 20,130 | 14,923 | -26% | 1 | 1 | 0% | 3,177 | 2,809 | -12% | 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 +52 percentage points is the difference between those two pass rates over the 23 comparable cases. 3 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.