Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate a Startup Canvas combining Product Strategy (9 sections) and Business Model (costs + revenue) for a new product. An alternative to BMC and Lean Canvas that separates strategy from business model. Use when launching a new product or evaluating a startup concept.
.claude/skills/phuryn-startup-canvas/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 66% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 71% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 87% | 0% |
Popular approaches like Business Model Canvas (Strategyzer) and Lean Canvas (Ash Maurya) mix strategy and business model into one artifact. The Startup Canvas (Paweł Huryn) separates them: 9 strategy sections from the Product Strategy Canvas + Cost Structure & Revenue Streams.
Why not Business Model Canvas?
Why not Lean Canvas?
When to use which:
You are a product strategist and startup advisor designing a Startup Canvas for $ARGUMENTS.
Your task is to create a comprehensive Startup Canvas that covers both the strategic choices and the business model for a new product.
1. Vision
2. Market Segments
3. Relative Costs
4. Value Proposition For each market segment:
5. Trade-offs
6. Key Metrics
7. Growth
8. Capabilities
9. Can't/Won't
10. Cost Structure
11. Revenue Streams
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 30,641 | 33,865 | +11% | 1 | 1 | 0% | 5,074 | 7,050 | +39% | 0 | 0 | — |
case-02 | fail→fail | 35,242 | 26,396 | -25% | 1 | 1 | 0% | 5,336 | 5,991 | +12% | 0 | 0 | — |
case-03 | fail→pass | 46,198 | 33,102 | -28% | 1 | 1 | 0% | 6,223 | 7,085 | +14% | 0 | 0 | — |
case-04 | pass→pass | 18,405 | 27,489 | +49% | 1 | 1 | 0% | 3,224 | 4,912 | +52% | 0 | 0 | — |
case-05 | pass→pass | 12,420 | 13,151 | +6% | 1 | 1 | 0% | 2,364 | 3,638 | +54% | 0 | 0 | — |
case-06 | pass→pass | 79,745 | 23,536 | -70% | 1 | 1 | 0% | 3,615 | 5,414 | +50% | 0 | 0 | — |
case-07 | fail→pass | 23,600 | 40,953 | +74% | 1 | 1 | 0% | 3,780 | 6,293 | +66% | 0 | 0 | — |
case-08 | pass→pass | 31,640 | 28,875 | -9% | 1 | 1 | 0% | 3,917 | 6,500 | +66% | 0 | 0 | — |
case-09 | fail→pass | 16,204 | 17,090 | +5% | 1 | 1 | 0% | 2,597 | 4,434 | +71% | 0 | 0 | — |
case-10 | fail→pass | 23,331 | 26,500 | +14% | 1 | 1 | 0% | 3,197 | 5,988 | +87% | 0 | 0 | — |
case-11 | pass→pass | 20,951 | 31,253 | +49% | 1 | 1 | 0% | 3,454 | 6,604 | +91% | 0 | 0 | — |
case-12 | fail→pass | 15,846 | 22,212 | +40% | 1 | 1 | 0% | 2,830 | 5,022 | +77% | 0 | 0 | — |
case-13 | fail→pass | 20,508 | 35,910 | +75% | 1 | 1 | 0% | 3,000 | 6,524 | +117% | 0 | 0 | — |
case-14 | fail→pass | 20,265 | 32,837 | +62% | 1 | 1 | 0% | 3,321 | 6,802 | +105% | 0 | 0 | — |
case-15 | fail→fail | 18,963 | 25,567 | +35% | 1 | 1 | 0% | 3,118 | 5,642 | +81% | 0 | 0 | — |
case-16 | pass→pass | 24,144 | 36,663 | +52% | 1 | 1 | 0% | 4,025 | 6,608 | +64% | 0 | 0 | — |
case-17 | fail→pass | 16,544 | 28,451 | +72% | 1 | 1 | 0% | 2,531 | 6,170 | +144% | 0 | 0 | — |
case-18 | fail→pass | 19,900 | 30,957 | +56% | 1 | 1 | 0% | 3,089 | 6,412 | +108% | 0 | 0 | — |
case-19 | pass→pass | 26,198 | 32,409 | +24% | 1 | 1 | 0% | 3,169 | 6,790 | +114% | 0 | 0 | — |
case-20 | fail→pass | 15,880 | 23,432 | +48% | 1 | 1 | 0% | 2,552 | 5,385 | +111% | 0 | 0 | — |
case-21 | fail→pass | 18,115 | 29,825 | +65% | 1 | 1 | 0% | 2,941 | 5,595 | +90% | 0 | 0 | — |
case-22 | fail→pass | 16,108 | 21,278 | +32% | 1 | 1 | 0% | 2,742 | 4,975 | +81% | 0 | 0 | — |
case-23 | fail→pass | 17,895 | 29,098 | +63% | 1 | 1 | 0% | 3,106 | 6,601 | +113% | 0 | 0 | — |
case-24 | fail→pass | 15,549 | 25,275 | +63% | 1 | 1 | 0% | 2,549 | 5,514 | +116% | 0 | 0 | — |
case-25 | fail→pass | 19,985 | 28,123 | +41% | 1 | 1 | 0% | 3,266 | 6,395 | +96% | 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. 25 cases were attempted. The headline lift of +64 percentage points is the difference between those two pass rates over the 25 comparable cases.
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.