Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Product ideation expert using Product Trio approach and Opportunity Solution Trees for both new and existing products.
.claude/skills/borghei-brainstorm-ideas/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 125% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 84% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 99% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -19% | 0% |
Structured product ideation for both new product creation and existing product enhancement. This skill combines the Product Trio approach (PM + Designer + Engineer perspectives) with Teresa Torres' Opportunity Solution Tree framework to generate, evaluate, and prioritize product ideas systematically.
Before ideating, confirm these inputs. If any is unknown or vague, ASK — do not assume:
Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.
Load the reference that matches the task — keep this file lean and pull detail on demand:
identify-assumptions/ to map and prioritize assumptions.brainstorm-experiments/ to design validation experiments for key assumptions.pre-mortem/ before committing to build, to surface hidden risks.In Scope: Structured ideation facilitation using Product Trio approach, Opportunity Solution Tree mapping, idea prioritization with weighted scoring, SCAMPER and HMW supplementary techniques, idea documentation with validation plans, integration with downstream discovery skills.
Out of Scope: Assumption testing and experiment design (hand off to brainstorm-experiments/ and identify-assumptions/), detailed product requirements (hand off to execution/create-prd/), market research and competitive analysis, financial modeling for ideas.
Limitations: Ideation quality is bounded by the diversity of perspectives in the room -- remote-only sessions may reduce creative energy. Scoring models provide structured comparison but are not objective truth; they encode the biases of the scorers. Opportunity Solution Trees require ongoing user research to populate -- they are not a substitute for customer interviews.
| Integration | Direction | What Flows | |-------------|-----------|------------| | identify-assumptions/ | Ideas -> Assumptions | Top 5 ideas feed into assumption mapping for risk assessment | | brainstorm-experiments/ | Ideas -> Experiments | Riskiest assumptions from ideas become experiment candidates | | pre-mortem/ | Ideas -> Risk | Selected ideas run through pre-mortem before build commitment | | execution/create-prd/ | Ideas -> PRD | Validated ideas become PRD inputs with problem statement and success metrics | | execution/brainstorm-okrs/ | OKRs -> Ideas | Team OKRs define the target outcomes that frame ideation sessions | | execution/prioritization-frameworks/ | Ideas -> Prioritization | Scored ideas feed into RICE or other frameworks for backlog ordering |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 28,838 | 39,922 | +38% | 1 | 1 | 0% | 4,512 | 7,234 | +60% | 0 | 0 | — |
case-01 | fail→fail | 25,048 | 26,293 | +5% | 1 | 1 | 0% | 3,960 | 5,152 | +30% | 0 | 0 | — |
case-03 | fail→fail | 33,846 | 37,677 | +11% | 1 | 1 | 0% | 5,493 | 7,249 | +32% | 0 | 0 | — |
case-04 | fail→pass | 22,408 | 34,012 | +52% | 1 | 1 | 0% | 3,207 | 7,223 | +125% | 0 | 0 | — |
case-05 | fail→pass | 19,189 | 24,692 | +29% | 1 | 1 | 0% | 2,743 | 5,057 | +84% | 0 | 0 | — |
case-06 | fail→pass | 20,113 | 24,536 | +22% | 1 | 1 | 0% | 2,990 | 4,915 | +64% | 0 | 0 | — |
case-07 | fail→pass | 13,310 | 19,792 | +49% | 1 | 1 | 0% | 2,043 | 4,067 | +99% | 0 | 0 | — |
case-18 | fail→pass | 18,361 | 7,438 | -59% | 1 | 1 | 0% | 2,720 | 2,196 | -19% | 0 | 0 | — |
case-08 | fail→pass | 18,368 | 14,218 | -23% | 1 | 1 | 0% | 2,906 | 3,092 | +6% | 0 | 0 | — |
case-09 | pass→pass | 11,424 | 23,121 | +102% | 1 | 1 | 0% | 1,834 | 4,708 | +157% | 0 | 0 | — |
case-10 | pass→pass | 20,446 | 26,939 | +32% | 1 | 1 | 0% | 2,696 | 4,914 | +82% | 0 | 0 | — |
case-11 | pass→pass | 12,172 | 18,041 | +48% | 1 | 1 | 0% | 1,828 | 3,871 | +112% | 0 | 0 | — |
case-12 | pass→pass | 14,793 | 23,771 | +61% | 1 | 1 | 0% | 1,963 | 4,446 | +126% | 0 | 0 | — |
case-13 | pass→pass | 10,565 | 25,481 | +141% | 1 | 1 | 0% | 1,490 | 4,867 | +227% | 0 | 0 | — |
case-14 | fail→fail | 26,778 | 26,528 | -1% | 1 | 1 | 0% | 4,186 | 5,532 | +32% | 0 | 0 | — |
case-15 | fail→fail | 35,342 | 15,513 | -56% | 1 | 1 | 0% | 6,187 | 3,571 | -42% | 0 | 0 | — |
case-16 | fail→pass | 23,555 | 15,442 | -34% | 1 | 1 | 0% | 4,111 | 3,261 | -21% | 0 | 0 | — |
case-17 | fail→pass | 22,107 | 8,648 | -61% | 1 | 1 | 0% | 3,453 | 2,341 | -32% | 0 | 0 | — |
case-19 | fail→pass | 23,146 | 7,668 | -67% | 1 | 1 | 0% | 3,405 | 2,362 | -31% | 0 | 0 | — |
case-20 | fail→pass | 10,637 | 18,917 | +78% | 1 | 1 | 0% | 1,440 | 3,991 | +177% | 0 | 0 | — |
case-21 | fail→pass | 13,007 | 5,229 | -60% | 1 | 1 | 0% | 1,884 | 1,832 | -3% | 0 | 0 | — |
case-22 | fail→fail | 9,056 | 12,301 | +36% | 1 | 1 | 0% | 1,642 | 2,902 | +77% | 0 | 0 | — |
case-23 | fail→pass | 15,401 | 25,402 | +65% | 1 | 1 | 0% | 2,162 | 4,771 | +121% | 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 +52 percentage points is the difference between those two pass rates over the 23 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.