Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Brainstorm feature ideas for a new product in initial discovery from PM, Designer, and Engineer perspectives. Use when starting product discovery for a new product, exploring features for a startup idea, or doing initial ideation.
.claude/skills/phuryn-brainstorm-ideas-new/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 83% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 22% | 0% |
Multi-perspective ideation for initial product discovery of a new product. Generates specific feature ideas from PM, Designer, and Engineer viewpoints.
You are supporting initial product discovery for a new product: $ARGUMENTS.
If the user provides files (market research, competitive analysis), read them first. Use web search to understand the market if needed.
Initial Discovery vs Continuous Discovery: Initial Discovery focuses on vision, business model, and market validation — you're testing whether the product should exist. Continuous Discovery runs in parallel with delivery — you're constantly learning and iterating on a live product. This skill is for initial discovery.
The user will describe their target segment, opportunity, and desired outcomes. Work through these steps:
Think step by step. Save substantial output as a markdown document.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 18,869 | 20,921 | +11% | 1 | 1 | 0% | 3,401 | 4,035 | +19% | 0 | 0 | — |
case-02 | pass→pass | 19,862 | 26,436 | +33% | 1 | 1 | 0% | 3,406 | 4,980 | +46% | 0 | 0 | — |
case-03 | pass→pass | 20,302 | 25,157 | +24% | 1 | 1 | 0% | 3,371 | 4,441 | +32% | 0 | 0 | — |
case-04 | pass→fail | 13,437 | 5,473 | -59% | 1 | 1 | 0% | 2,208 | 1,376 | -38% | 0 | 0 | — |
case-05 | pass→pass | 17,480 | 26,044 | +49% | 1 | 1 | 0% | 2,893 | 4,791 | +66% | 0 | 0 | — |
case-06 | fail→pass | 16,873 | 28,703 | +70% | 1 | 1 | 0% | 2,889 | 5,296 | +83% | 0 | 0 | — |
case-07 | pass→pass | 12,011 | 18,249 | +52% | 1 | 1 | 0% | 1,944 | 3,328 | +71% | 0 | 0 | — |
case-08 | pass→pass | 17,426 | 24,109 | +38% | 1 | 1 | 0% | 2,680 | 4,345 | +62% | 0 | 0 | — |
case-09 | pass→pass | 17,979 | 27,260 | +52% | 1 | 1 | 0% | 2,909 | 5,183 | +78% | 0 | 0 | — |
case-10 | pass→pass | 31,726 | 13,228 | -58% | 1 | 1 | 0% | 2,127 | 2,543 | +20% | 0 | 0 | — |
case-11 | fail→pass | 28,370 | 35,431 | +25% | 1 | 1 | 0% | 4,681 | 5,294 | +13% | 0 | 0 | — |
case-12 | pass→pass | 16,552 | 25,160 | +52% | 1 | 1 | 0% | 2,504 | 4,616 | +84% | 0 | 0 | — |
case-13 | fail→pass | 17,600 | 22,678 | +29% | 1 | 1 | 0% | 2,783 | 4,218 | +52% | 0 | 0 | — |
case-14 | fail→pass | 24,175 | 57,760 | +139% | 1 | 1 | 0% | 3,426 | 4,196 | +22% | 0 | 0 | — |
case-15 | pass→pass | 10,383 | 20,602 | +98% | 1 | 1 | 0% | 1,674 | 3,855 | +130% | 0 | 0 | — |
case-16 | pass→pass | 15,683 | 18,572 | +18% | 1 | 1 | 0% | 2,396 | 3,420 | +43% | 0 | 0 | — |
case-17 | pass→pass | 14,013 | 27,489 | +96% | 1 | 1 | 0% | 2,313 | 3,645 | +58% | 0 | 0 | — |
case-18 | pass→pass | 13,615 | 28,506 | +109% | 1 | 1 | 0% | 2,133 | 3,946 | +85% | 0 | 0 | — |
case-19 | fail→pass | 16,284 | 27,293 | +68% | 1 | 1 | 0% | 2,523 | 3,728 | +48% | 0 | 0 | — |
case-20 | pass→pass | 9,631 | 24,164 | +151% | 1 | 1 | 0% | 1,561 | 4,403 | +182% | 0 | 0 | — |
case-21 | pass→pass | 14,686 | 39,018 | +166% | 1 | 1 | 0% | 2,402 | 5,019 | +109% | 0 | 0 | — |
case-22 | fail→pass | 17,200 | 23,081 | +34% | 1 | 1 | 0% | 2,775 | 4,641 | +67% | 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. The headline lift of +27 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.