Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Full 8-stage Gordon-Prince excursion process. Deliberate departure from the problem into unrelated domains, then force-fit discoveries back.
.claude/skills/yogsoth-ai-excursion-method/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -12% | 0% |
Full 8-stage Gordon-Prince excursion process — deliberate departure from the problem into unrelated domains, then force-fit discoveries back to generate breakthrough solutions.
| Resource | Target | Current | % | |----------|--------|---------|---| | web-search | 30 | 0 | 0% | | web-research | 10 | 0 | 0% | | paper-overview | 25 | 0 | 0% | | paper-search | 15 | 0 | 0% | | paper-research | 8 | 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 | |--------|------| | excursion-orchestration | Orchestrate departure → force-fit → springboard | | compressed-conflict | Generate compressed conflicts during excursion | | analogy-extraction | Extract analogies from excursion domain |
| SOP | Role | |-----|------| | excursion-departure | Leave problem, explore unrelated domain | | direct-analogy-generation | Find analogies in excursion domain | | personal-identification | Embody elements in excursion domain | | symbolic-compression | Compress excursion findings into oxymorons | | force-fit | Force-fit discoveries back to problem | | springboard-launch | Convert force-fitted ideas into solutions | | synectics-synthesis | Synthesize full excursion report |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | compressed-conflict | Generate compressed conflicts (oxymorons) from problem contradictions and extract concrete idea directions from the symbolic tension. | | excursion-orchestration | Orchestrate the excursion sequence — departure into unrelated domain, force-fit discoveries back to problem, launch springboard ideas. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | direct-analogy-generation | Find direct analogies from nature/tech/society that share structural properties with the problem. Produces analogy list with structural mappings. | | excursion-departure | Leave the problem entirely and explore an unrelated domain. Produces excursion domain discoveries for later force-fitting. | | force-fit | Force-fit excursion discoveries back to the original problem. Deliberately create connections between unrelated findings and the challenge. | | personal-identification | First-person empathic identification with a system or component. Produces experience description and design insights from embodiment. | | springboard-launch | Convert analogy insights into concrete feasible solutions. Transform abstract connections into actionable mechanisms. | | symbolic-compression | Compress problem contradiction into 2-3 word oxymoron. Produces oxymorons with interpretation directions for each. | | synectics-synthesis | Synthesize all synectics outputs into a structured idea report. Combines results from all analogy types and excursion processes. |
<!-- 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 | pass→pass | 10,092 | 6,527 | -35% | 1 | 1 | 0% | 1,737 | 1,875 | +8% | 0 | 0 | — |
case-02 | fail→pass | 11,455 | 8,994 | -21% | 1 | 1 | 0% | 1,839 | 2,282 | +24% | 0 | 0 | — |
case-03 | fail→pass | 26,453 | 2,858 | -89% | 1 | 1 | 0% | 1,981 | 1,349 | -32% | 0 | 0 | — |
case-04 | fail→pass | 19,779 | 1,694 | -91% | 1 | 1 | 0% | 1,452 | 1,149 | -21% | 0 | 0 | — |
case-05 | fail→pass | 9,563 | 2,093 | -78% | 1 | 1 | 0% | 1,375 | 1,185 | -14% | 0 | 0 | — |
case-06 | pass→pass | 6,702 | 7,393 | +10% | 1 | 1 | 0% | 1,123 | 2,027 | +80% | 0 | 0 | — |
case-07 | fail→pass | 9,849 | 2,668 | -73% | 1 | 1 | 0% | 1,429 | 1,259 | -12% | 0 | 0 | — |
case-08 | pass→pass | 4,963 | 7,342 | +48% | 1 | 1 | 0% | 848 | 2,058 | +143% | 0 | 0 | — |
case-09 | pass→pass | 10,075 | 8,893 | -12% | 1 | 1 | 0% | 1,469 | 2,127 | +45% | 0 | 0 | — |
case-10 | fail→pass | 15,870 | 12,392 | -22% | 1 | 1 | 0% | 2,538 | 3,000 | +18% | 0 | 0 | — |
case-11 | pass→pass | 8,473 | 12,243 | +44% | 1 | 1 | 0% | 1,208 | 2,617 | +117% | 0 | 0 | — |
case-12 | pass→pass | 10,356 | 12,119 | +17% | 1 | 1 | 0% | 1,506 | 2,747 | +82% | 0 | 0 | — |
case-13 | pass→pass | 8,165 | 5,008 | -39% | 1 | 1 | 0% | 1,254 | 1,648 | +31% | 0 | 0 | — |
case-14 | pass→pass | 13,021 | 7,912 | -39% | 1 | 1 | 0% | 1,951 | 2,060 | +6% | 0 | 0 | — |
case-15 | pass→pass | 7,155 | 11,318 | +58% | 1 | 1 | 0% | 1,003 | 2,451 | +144% | 0 | 0 | — |
case-16 | pass→pass | 8,656 | 2,444 | -72% | 1 | 1 | 0% | 1,330 | 1,220 | -8% | 0 | 0 | — |
case-17 | fail→pass | 9,019 | 8,881 | -2% | 1 | 1 | 0% | 1,340 | 2,261 | +69% | 0 | 0 | — |
case-18 | pass→pass | 9,050 | 7,592 | -16% | 1 | 1 | 0% | 1,347 | 1,964 | +46% | 0 | 0 | — |
case-19 | pass→pass | 13,188 | 30,892 | +134% | 1 | 1 | 0% | 2,074 | 5,553 | +168% | 0 | 0 | — |
case-20 | pass→pass | 23,511 | 42,656 | +81% | 1 | 1 | 0% | 3,319 | 7,038 | +112% | 0 | 0 | — |
case-21 | pass→pass | 17,472 | 29,656 | +70% | 1 | 1 | 0% | 2,831 | 5,542 | +96% | 0 | 0 | — |
case-22 | fail→pass | 14,252 | 4,504 | -68% | 1 | 1 | 0% | 891 | 1,748 | +96% | 0 | 0 | — |
case-23 | fail→pass | 16,118 | 2,928 | -82% | 1 | 1 | 0% | 2,730 | 1,287 | -53% | 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, and 20 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 +39 percentage points is the difference between those two pass rates over the 20 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.