Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Complete DBA process: problem reframe → source search → map → transfer → adapt. Full Design-by-Analogy methodology for systematic analogical design.
.claude/skills/yogsoth-ai-design-by-analogy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 278% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 13% | 0% |
Complete Design-by-Analogy (DBA) process following the full methodology: problem reframe → source search → structural mapping → principle transfer → target adaptation.
| Resource | Target | Current | % | |----------|--------|---------|---| | web-search | 30 | 0 | 0% | | web-research | 10 | 0 | 0% | | paper-overview | 30 | 0 | 0% | | paper-search | 20 | 0 | 0% | | paper-research | 10 | 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 | |--------|------| | analogy-extraction | Core tactic — full analogy extraction pipeline | | domain-divergence | Maximize source domain diversity | | bridge-validation | Validate analogy quality before transfer |
| SOP | Role | |-----|------| | domain-scanning | Search for analogous source domains | | abstraction-extraction | Extract abstract design principles from source | | structural-mapping | Map source→target structural correspondences | | analogy-quality-assessment | Assess analogy depth and transfer viability | | transfer-adaptation | Adapt transferred principle to target constraints | | cross-domain-synthesis | Synthesize DBA outputs into design proposals |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | bridge-validation | Validate analogy depth and transfer viability. Ensures only deep structural analogies (not surface-level similarities) proceed to transfer. | | domain-divergence | Scan and select maximally diverse source domains. Ensures creative search covers genuinely unrelated fields with high transfer potential. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | abstraction-extraction | Extract abstract principles from concrete domain cases. Strips domain-specific details to reveal transferable mechanisms. | | analogy-quality-assessment | Assess analogy depth (surface/structural/systemic). Determines whether an analogy warrants transfer investment. | | cross-domain-synthesis | Synthesize all cross-domain findings into a structured idea report. Integrates outputs from all strategies and SOPs. | | structural-mapping | Map source→target structural correspondences. Identifies corresponding, missing, and extra elements between domains. | | transfer-adaptation | Adapt transferred principle to target problem constraints. Produces concrete adapted solutions from abstract principles. |
<!-- 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-12 | pass→pass | 15,545 | 37,045 | +138% | 1 | 1 | 0% | 2,309 | 5,909 | +156% | 0 | 0 | — |
case-11 | fail→fail | 25,849 | 46,125 | +78% | 1 | 1 | 0% | 3,809 | 6,976 | +83% | 0 | 0 | — |
case-01 | fail→pass | 101,405 | 47,267 | -53% | 1 | 1 | 0% | 6,218 | 7,007 | +13% | 0 | 0 | — |
case-02 | fail→fail | 40,747 | 38,935 | -4% | 1 | 1 | 0% | 6,216 | 7,004 | +13% | 0 | 0 | — |
case-03 | fail→fail | 36,427 | 37,108 | +2% | 1 | 1 | 0% | 6,208 | 6,997 | +13% | 0 | 0 | — |
case-04 | fail→pass | 26,487 | 30,683 | +16% | 1 | 1 | 0% | 6,210 | 6,999 | +13% | 0 | 0 | — |
case-05 | fail→pass | 9,394 | 29,323 | +212% | 1 | 1 | 0% | 1,846 | 6,985 | +278% | 0 | 0 | — |
case-06 | pass→fail | 27,133 | 26,273 | -3% | 1 | 1 | 0% | 5,767 | 1,370 | -76% | 0 | 0 | — |
case-07 | pass→fail | 16,478 | 10,221 | -38% | 1 | 1 | 0% | 2,589 | 1,427 | -45% | 0 | 0 | — |
case-08 | fail→fail | 11,218 | 12,012 | +7% | 1 | 1 | 0% | 1,623 | 1,422 | -12% | 0 | 0 | — |
case-09 | fail→fail | 27,413 | 10,427 | -62% | 1 | 1 | 0% | 4,319 | 1,461 | -66% | 0 | 0 | — |
case-10 | pass→fail | 18,408 | 8,496 | -54% | 1 | 1 | 0% | 2,653 | 1,293 | -51% | 0 | 0 | — |
case-13 | fail→fail | 41,720 | 8,683 | -79% | 1 | 1 | 0% | 6,179 | 1,453 | -76% | 0 | 0 | — |
case-14 | fail→fail | 22,313 | 50,376 | +126% | 1 | 1 | 0% | 3,314 | 8,050 | +143% | 0 | 0 | — |
case-15 | fail→pass | 28,803 | 37,911 | +32% | 1 | 1 | 0% | 4,258 | 6,974 | +64% | 0 | 0 | — |
case-16 | fail→pass | 39,912 | 40,352 | +1% | 1 | 1 | 0% | 6,187 | 6,976 | +13% | 0 | 0 | — |
case-17 | fail→pass | 26,750 | 34,718 | +30% | 1 | 1 | 0% | 4,348 | 6,963 | +60% | 0 | 0 | — |
case-18 | fail→pass | 24,603 | 38,319 | +56% | 1 | 1 | 0% | 3,955 | 6,962 | +76% | 0 | 0 | — |
case-19 | fail→pass | 19,280 | 39,595 | +105% | 1 | 1 | 0% | 2,831 | 6,960 | +146% | 0 | 0 | — |
case-20 | fail→fail | 24,398 | 10,553 | -57% | 1 | 1 | 0% | 3,821 | 1,419 | -63% | 0 | 0 | — |
case-21 | pass→fail | 21,359 | 9,420 | -56% | 1 | 1 | 0% | 3,372 | 1,298 | -62% | 0 | 0 | — |
case-22 | fail→pass | 28,369 | 38,308 | +35% | 1 | 1 | 0% | 4,319 | 6,962 | +61% | 0 | 0 | — |
case-23 | fail→pass | 18,060 | 40,900 | +126% | 1 | 1 | 0% | 2,764 | 6,961 | +152% | 0 | 0 | — |
case-24 | fail→fail | 30,138 | 9,969 | -67% | 1 | 1 | 0% | 4,710 | 1,301 | -72% | 0 | 0 | — |
case-25 | fail→fail | 19,239 | 27,772 | +44% | 1 | 1 | 0% | 2,928 | 1,105 | -62% | 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. 25 cases were attempted, and 14 counted toward the lift figure. The other 11 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 +24 percentage points is the difference between those two pass rates over the 14 comparable cases. 5 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.