MaybeMonad Brainstorm Prototype
Use this skill when the user needs to see or compare possibilities before choosing a direction. The output should make implicit criteria visible early, while changes are still cheap.
Workflow
- Identify the decision surface:
- Product strategy, UI layout, data view, interaction model, technical approach, workflow, copy, or visual style.
- Name the comparison axes:
- Cost, ambition, risk, reversibility, UX density, implementation path, maintenance burden, time to validate, or taste direction.
- Generate meaningfully different options:
- Prefer 3-5 distinct directions over small variations.
- Label each option by the intuition it tests.
- Prototype only as much as needed:
- Use static HTML, fake data, sketches, local mock state, or a minimal isolated component when that helps the user react.
- Do not wire backend routes, migrations, durable state, payments, auth, or production side effects unless explicitly asked.
- Explain what each option teaches and what implementation path it implies.
- Ask the user to react to tradeoffs, not just pick a favorite.
Output Shape
For brainstorming:
- Options table: direction, what it optimizes, tradeoffs, implementation implications, unknowns it tests.
- Recommendation: one default and one backup.
- Next probe: the cheapest prototype or reference check.
For prototyping:
- Prototype scope: what is real and what is fake.
- How to view it, if a file or local server is created.
- Reaction prompts: 3-5 concrete things the user should judge.
Rules
- Make options genuinely different. Do not pad with cosmetic variants.
- Preserve exact user copy, screenshots, and semantics.
- Prefer cheap throwaway artifacts over changing the real app when the user's criteria are still forming.
- If prototyping would take longer than a narrow code change, say why and offer the smallest useful alternative.