Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Synthesize all assessments into a feasibility matrix, recommendation, and risk summary.
.claude/skills/yogsoth-ai-feasibility-synthesis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 70% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 201% | 0% |
Combine all feasibility assessment outputs — readiness scores, constraint analyses, resource estimates, and gate verdicts — into a final feasibility matrix with clear recommendation and risk summary.
Spawns a subagent that:
Final synthesis requires integrating diverse assessment types into a coherent whole. A dedicated subagent can weigh conflicting signals and produce a balanced recommendation without anchoring to any single assessment.
Output MUST include: feasibility matrix, clear recommendation (proceed/defer/abandon), and risk summary with top 3 risks. Reject if recommendation is not supported by the matrix data.
<!-- 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-11 | fail→pass | 27,757 | 19,456 | -30% | 1 | 1 | 0% | 4,553 | 3,429 | -25% | 0 | 0 | — |
case-18 | fail→pass | 17,934 | 26,348 | +47% | 1 | 1 | 0% | 2,901 | 4,929 | +70% | 0 | 0 | — |
case-09 | pass→pass | 14,234 | 12,337 | -13% | 1 | 1 | 0% | 2,132 | 2,273 | +7% | 0 | 0 | — |
case-10 | fail→pass | 13,751 | 15,875 | +15% | 1 | 1 | 0% | 2,414 | 2,900 | +20% | 0 | 0 | — |
case-01 | fail→pass | 14,321 | 20,881 | +46% | 1 | 1 | 0% | 2,244 | 2,933 | +31% | 0 | 0 | — |
case-02 | fail→fail | 2,959 | 19,131 | +547% | 1 | 1 | 0% | 479 | 3,540 | +639% | 0 | 0 | — |
case-03 | pass→pass | 18,612 | 21,699 | +17% | 1 | 1 | 0% | 3,123 | 3,943 | +26% | 0 | 0 | — |
case-04 | pass→fail | 20,506 | 33,088 | +61% | 1 | 1 | 0% | 3,586 | 6,441 | +80% | 0 | 0 | — |
case-05 | pass→pass | 10,074 | 18,264 | +81% | 1 | 1 | 0% | 1,676 | 3,478 | +108% | 0 | 0 | — |
case-06 | fail→fail | 6,400 | 16,933 | +165% | 1 | 1 | 0% | 1,067 | 3,288 | +208% | 0 | 0 | — |
case-07 | fail→fail | 21,353 | 22,493 | +5% | 1 | 1 | 0% | 3,102 | 3,611 | +16% | 0 | 0 | — |
case-08 | fail→fail | 17,047 | 14,976 | -12% | 1 | 1 | 0% | 2,812 | 2,785 | -1% | 0 | 0 | — |
case-12 | fail→pass | 6,542 | 13,729 | +110% | 1 | 1 | 0% | 884 | 2,657 | +201% | 0 | 0 | — |
case-13 | pass→pass | 12,845 | 14,121 | +10% | 1 | 1 | 0% | 1,997 | 2,545 | +27% | 0 | 0 | — |
case-14 | fail→fail | 10,548 | 13,558 | +29% | 1 | 1 | 0% | 1,597 | 2,438 | +53% | 0 | 0 | — |
case-15 | fail→fail | 13,194 | 15,898 | +20% | 1 | 1 | 0% | 2,055 | 2,916 | +42% | 0 | 0 | — |
case-16 | fail→fail | 19,139 | 18,343 | -4% | 1 | 1 | 0% | 3,117 | 3,267 | +5% | 0 | 0 | — |
case-17 | fail→pass | 10,456 | 11,439 | +9% | 1 | 1 | 0% | 1,726 | 2,163 | +25% | 0 | 0 | — |
case-19 | fail→pass | 11,305 | 13,286 | +18% | 1 | 1 | 0% | 1,587 | 2,456 | +55% | 0 | 0 | — |
case-20 | pass→pass | 13,522 | 20,870 | +54% | 1 | 1 | 0% | 2,282 | 3,914 | +72% | 0 | 0 | — |
case-21 | fail→fail | 14,136 | 19,511 | +38% | 1 | 1 | 0% | 2,299 | 3,608 | +57% | 0 | 0 | — |
case-22 | fail→fail | 13,684 | 7,308 | -47% | 1 | 1 | 0% | 2,272 | 619 | -73% | 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 21 counted toward the lift figure. The other 1 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 +27 percentage points is the difference between those two pass rates over the 21 comparable cases. 2 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.