Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →13 research methodology skills. Trigger: study design, methodology selection, scientific reasoning, mentoring. Design: rigorous methods frameworks covering qualitative, quantitative, and mixed approaches.
.claude/skills/brycewang-stanford-methodology-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -50% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -66% | 0% |
Select the skill matching the user's need, then read its SKILL.md.
| Skill | Description | |-------|-------------| | action-research-guide | Design and conduct action research and participatory studies | | claude-scientific-guide | Ready-to-use agent skills for scientific research and engineering | | experimental-design-guide | Design rigorous experiments using DOE, factorial designs, and response surfaces | | grad-school-guide | Practical advice for thriving in PhD programs and academic research | | grounded-theory-guide | Apply grounded theory methodology to develop theory from data | | mixed-methods-guide | Guide to designing and conducting mixed methods research | | osf-api-guide | Access Open Science Framework for preregistrations, preprints, and data | | parsifal-slr-guide | Plan and manage systematic literature reviews with Parsifal platform | | qualitative-research-guide | Design and conduct qualitative research using grounded theory and case studies | | research-paper-kb | Build a persistent cross-session knowledge base from academic papers | | research-town-guide | Simulate human research communities with multi-agent AI collaboration | | scientify-idea-generation | Generate research ideas from collected papers with gap analysis | | slr-automation-guide | Tools and pipelines for automating systematic literature reviews |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | pass→pass | 16,221 | 4,908 | -70% | 1 | 1 | 0% | 2,463 | 1,326 | -46% | 0 | 0 | — |
case-17 | pass→pass | 8,929 | 10,543 | +18% | 1 | 1 | 0% | 1,517 | 1,456 | -4% | 0 | 0 | — |
case-01 | fail→fail | 18,161 | 65,661 | +262% | 1 | 1 | 0% | 3,183 | 983 | -69% | 0 | 0 | — |
case-02 | fail→fail | 20,693 | 34,429 | +66% | 1 | 1 | 0% | 2,876 | 887 | -69% | 0 | 0 | — |
case-03 | pass→fail | 14,503 | 6,595 | -55% | 1 | 1 | 0% | 2,434 | 999 | -59% | 0 | 0 | — |
case-04 | pass→fail | 19,297 | 36,304 | +88% | 1 | 1 | 0% | 2,880 | 966 | -66% | 0 | 0 | — |
case-10 | pass→pass | 10,331 | 4,372 | -58% | 1 | 1 | 0% | 1,684 | 1,222 | -27% | 0 | 0 | — |
case-05 | pass→pass | 14,104 | 4,704 | -67% | 1 | 1 | 0% | 2,099 | 1,189 | -43% | 0 | 0 | — |
case-06 | fail→pass | 17,052 | 6,763 | -60% | 1 | 1 | 0% | 2,833 | 1,603 | -43% | 0 | 0 | — |
case-07 | fail→pass | 45,546 | 6,740 | -85% | 1 | 1 | 0% | 2,584 | 1,236 | -52% | 0 | 0 | — |
case-08 | fail→pass | 14,858 | 3,308 | -78% | 1 | 1 | 0% | 2,161 | 1,089 | -50% | 0 | 0 | — |
case-09 | fail→pass | 46,237 | 4,944 | -89% | 1 | 1 | 0% | 2,364 | 1,261 | -47% | 0 | 0 | — |
case-12 | fail→pass | 17,437 | 2,873 | -84% | 1 | 1 | 0% | 3,039 | 1,023 | -66% | 0 | 0 | — |
case-13 | pass→fail | 8,483 | 7,364 | -13% | 1 | 1 | 0% | 1,574 | 1,256 | -20% | 0 | 0 | — |
case-14 | fail→pass | 11,608 | 4,887 | -58% | 1 | 1 | 0% | 1,784 | 1,219 | -32% | 0 | 0 | — |
case-15 | fail→pass | 9,231 | 3,850 | -58% | 1 | 1 | 0% | 1,553 | 1,180 | -24% | 0 | 0 | — |
case-16 | pass→fail | 13,381 | 5,115 | -62% | 1 | 1 | 0% | 2,178 | 943 | -57% | 0 | 0 | — |
case-18 | pass→pass | 15,829 | 7,941 | -50% | 1 | 1 | 0% | 2,303 | 1,151 | -50% | 0 | 0 | — |
case-19 | fail→pass | 16,253 | 5,155 | -68% | 1 | 1 | 0% | 2,585 | 1,213 | -53% | 0 | 0 | — |
case-20 | fail→pass | 17,361 | 13,554 | -22% | 1 | 1 | 0% | 2,567 | 2,449 | -5% | 0 | 0 | — |
case-21 | pass→fail | 11,435 | 5,952 | -48% | 1 | 1 | 0% | 1,653 | 767 | -54% | 0 | 0 | — |
case-22 | pass→pass | 8,973 | 7,764 | -13% | 1 | 1 | 0% | 1,465 | 1,751 | +20% | 0 | 0 | — |
case-23 | pass→pass | 4,057 | 9,322 | +130% | 1 | 1 | 0% | 906 | 1,900 | +110% | 0 | 0 | — |
case-24 | pass→pass | 17,194 | 25,480 | +48% | 1 | 1 | 0% | 2,713 | 2,531 | -7% | 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. 24 cases were attempted, and 17 counted toward the lift figure. The other 7 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 +17 percentage points is the difference between those two pass rates over the 17 comparable cases. 6 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.