Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →SOP: Match the most suitable RQ framework based on research type
.claude/skills/yogsoth-ai-framework-matching/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -19% | 0% |
Match the most suitable RQ framework based on research type.
<HARD-GATE> Input must contain: research direction/hypothesis + research type (or enough information to infer the research type). </HARD-GATE>
| Framework | Applicable research type | Core components | Strengths | Limitations | |------|------------|---------|------|------| | PICO/PICOTS | Quantitative/intervention | P-I-C-O(-T-S) | Clear structure, widely recognized | Not suitable for qualitative research | | SPIDER | Qualitative | S-PI-D-E-R | Designed specifically for qualitative | Not suitable for intervention research | | SPICE | Evaluation | S-P-I-C-E | Evaluation-oriented | Narrower scope | | ECLIPSE | Mixed methods | E-C-L-I-P-S-E | Comprehensive, suitable for complex research | Many components, may be excessive |
{
recommended: "PICO",
rationale: "...",
candidates: [
{ framework: "PICO", fit_score: 9, reason: "..." },
{ framework: "SPIDER", fit_score: 4, reason: "..." }
]
}| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 7,192 | 14,293 | +99% | 1 | 1 | 0% | 1,387 | 1,603 | +16% | 0 | 0 | — |
case-02 | fail→pass | 9,404 | 5,627 | -40% | 1 | 1 | 0% | 1,597 | 1,482 | -7% | 0 | 0 | — |
case-08 | fail→pass | 10,133 | 8,402 | -17% | 1 | 1 | 0% | 2,011 | 1,860 | -8% | 0 | 0 | — |
case-03 | fail→pass | 12,479 | 7,491 | -40% | 1 | 1 | 0% | 2,223 | 1,798 | -19% | 0 | 0 | — |
case-04 | fail→pass | 11,352 | 6,707 | -41% | 1 | 1 | 0% | 1,984 | 1,598 | -19% | 0 | 0 | — |
case-05 | fail→pass | 8,124 | 5,960 | -27% | 1 | 1 | 0% | 1,507 | 1,478 | -2% | 0 | 0 | — |
case-06 | fail→pass | 8,716 | 5,160 | -41% | 1 | 1 | 0% | 1,501 | 1,458 | -3% | 0 | 0 | — |
case-07 | fail→pass | 12,158 | 5,993 | -51% | 1 | 1 | 0% | 2,153 | 1,492 | -31% | 0 | 0 | — |
case-09 | fail→pass | 12,425 | 5,146 | -59% | 1 | 1 | 0% | 2,095 | 1,274 | -39% | 0 | 0 | — |
case-10 | fail→pass | 7,089 | 5,415 | -24% | 1 | 1 | 0% | 1,167 | 1,400 | +20% | 0 | 0 | — |
case-11 | fail→pass | 9,120 | 7,624 | -16% | 1 | 1 | 0% | 1,367 | 1,681 | +23% | 0 | 0 | — |
case-12 | fail→pass | 9,177 | 8,057 | -12% | 1 | 1 | 0% | 1,493 | 1,772 | +19% | 0 | 0 | — |
case-17 | fail→pass | 3,886 | 4,830 | +24% | 1 | 1 | 0% | 702 | 1,330 | +89% | 0 | 0 | — |
case-13 | fail→pass | 9,692 | 6,109 | -37% | 1 | 1 | 0% | 1,857 | 1,560 | -16% | 0 | 0 | — |
case-14 | fail→pass | 9,779 | 5,142 | -47% | 1 | 1 | 0% | 1,543 | 1,303 | -16% | 0 | 0 | — |
case-15 | fail→pass | 9,329 | 6,924 | -26% | 1 | 1 | 0% | 1,636 | 1,648 | +1% | 0 | 0 | — |
case-16 | fail→pass | 8,188 | 5,906 | -28% | 1 | 1 | 0% | 1,397 | 1,416 | +1% | 0 | 0 | — |
case-18 | fail→pass | 6,501 | 5,598 | -14% | 1 | 1 | 0% | 1,141 | 1,406 | +23% | 0 | 0 | — |
case-19 | fail→pass | 8,790 | 6,149 | -30% | 1 | 1 | 0% | 1,545 | 1,505 | -3% | 0 | 0 | — |
case-20 | fail→pass | 10,872 | 6,666 | -39% | 1 | 1 | 0% | 1,928 | 1,585 | -18% | 0 | 0 | — |
case-21 | pass→fail | 7,871 | 9,615 | +22% | 1 | 1 | 0% | 1,596 | 2,369 | +48% | 0 | 0 | — |
case-22 | pass→fail | 5,481 | 8,566 | +56% | 1 | 1 | 0% | 1,136 | 2,020 | +78% | 0 | 0 | — |
case-23 | pass→fail | 14,755 | 8,964 | -39% | 1 | 1 | 0% | 2,737 | 2,091 | -24% | 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 +74 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.