Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Transform rough ideas into solid designs through structured questioning
.claude/skills/dicklesworthstone-brainstorming/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 48% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -5% | 0% |
Turn vague ideas into actionable designs through Socratic questioning.
> "I'm using the Brainstorming skill to refine your idea into a design."
[Idea] → Phase 1: Understand → Phase 2: Explore → Phase 3: Design → [Plan]One question at a time. Prefer multiple choice.
Questions to ask:
Before asking: Check working directory for existing context.
Present 2-3 approaches:
markdown### Approach A: [Name] - Architecture: [how it works] - Pros: [benefits] - Cons: [tradeoffs] - Complexity: [low/medium/high] ### Approach B: [Name] ...
Ask: "Which approach resonates? Or should we explore others?"
Present design in 200-300 word chunks:
After each chunk: "Does this look right?"
When design approved:
> "Ready to create the implementation plan?"
If yes → Use writing-plans skill
It's okay to revisit earlier phases:
Don't force linear progress.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | fail→pass | 14,250 | 5,964 | -58% | 1 | 1 | 0% | 2,572 | 1,378 | -46% | 0 | 0 | — |
case-01 | fail→pass | 10,515 | 9,283 | -12% | 1 | 1 | 0% | 1,474 | 1,121 | -24% | 0 | 0 | — |
case-02 | fail→pass | 11,643 | 4,367 | -62% | 1 | 1 | 0% | 1,837 | 1,299 | -29% | 0 | 0 | — |
case-03 | fail→pass | 7,491 | 6,052 | -19% | 1 | 1 | 0% | 986 | 1,459 | +48% | 0 | 0 | — |
case-04 | fail→fail | 13,445 | 7,415 | -45% | 1 | 1 | 0% | 1,926 | 1,509 | -22% | 0 | 0 | — |
case-05 | fail→pass | 9,459 | 5,692 | -40% | 1 | 1 | 0% | 1,493 | 1,413 | -5% | 0 | 0 | — |
case-07 | fail→fail | 18,590 | 9,667 | -48% | 1 | 1 | 0% | 2,926 | 1,833 | -37% | 0 | 0 | — |
case-08 | fail→fail | 16,889 | 9,916 | -41% | 1 | 1 | 0% | 2,348 | 2,007 | -15% | 0 | 0 | — |
case-09 | fail→fail | 41,089 | 15,970 | -61% | 1 | 1 | 0% | 8,227 | 2,761 | -66% | 0 | 0 | — |
case-10 | fail→fail | 26,508 | 9,549 | -64% | 1 | 1 | 0% | 5,335 | 2,037 | -62% | 0 | 0 | — |
case-11 | pass→pass | 10,473 | 9,030 | -14% | 1 | 1 | 0% | 1,488 | 1,910 | +28% | 0 | 0 | — |
case-12 | fail→pass | 8,714 | 6,154 | -29% | 1 | 1 | 0% | 1,326 | 1,472 | +11% | 0 | 0 | — |
case-13 | pass→pass | 5,323 | 1,848 | -65% | 1 | 1 | 0% | 817 | 692 | -15% | 0 | 0 | — |
case-14 | pass→pass | 9,894 | 5,135 | -48% | 1 | 1 | 0% | 1,644 | 1,179 | -28% | 0 | 0 | — |
case-15 | fail→pass | 21,048 | 5,652 | -73% | 1 | 1 | 0% | 2,947 | 1,251 | -58% | 0 | 0 | — |
case-16 | fail→pass | 14,029 | 6,216 | -56% | 1 | 1 | 0% | 1,785 | 1,468 | -18% | 0 | 0 | — |
case-17 | fail→pass | 12,178 | 6,015 | -51% | 1 | 1 | 0% | 1,989 | 1,355 | -32% | 0 | 0 | — |
case-18 | pass→pass | 13,599 | 7,302 | -46% | 1 | 1 | 0% | 1,966 | 1,487 | -24% | 0 | 0 | — |
case-19 | pass→fail | 5,081 | 2,800 | -45% | 1 | 1 | 0% | 773 | 830 | +7% | 0 | 0 | — |
case-20 | fail→fail | 17,590 | 8,864 | -50% | 1 | 1 | 0% | 2,512 | 1,751 | -30% | 0 | 0 | — |
case-21 | fail→fail | 44,178 | 6,434 | -85% | 1 | 1 | 0% | 7,460 | 1,443 | -81% | 0 | 0 | — |
case-22 | pass→pass | 34,610 | 31,396 | -9% | 1 | 1 | 0% | 6,763 | 5,718 | -15% | 0 | 0 | — |
case-23 | pass→fail | 19,683 | 6,806 | -65% | 1 | 1 | 0% | 4,671 | 1,603 | -66% | 0 | 0 | — |
case-24 | pass→fail | 21,502 | 10,010 | -53% | 1 | 1 | 0% | 3,420 | 2,125 | -38% | 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. The headline lift of +25 percentage points is the difference between those two pass rates over the 24 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.