Startup MVP Development Skill

Core principle

The MVP's job is to test the riskiest assumption as cheaply as possible — not to be a smaller version of the final product. Before building anything, identify: what is the one thing that, if false, kills this idea? Build only enough to test that.

Scoping framework

1. List every feature the "full product" would need.
2. Tag each: core (the thing being tested), supporting (needed for core to make sense), nice-to-have (defer), fake-able (can be manual/hardcoded for now — e.g. a Stripe payment link instead of full billing, a Google Form instead of a settings page, manual email instead of an automated notification system).
3. Build only core + minimal supporting. Aggressively fake or cut everything else — a founder manually onboarding the first 20 users by hand is a valid MVP strategy.

Stack choices optimized for speed-to-validate

• Next.js + a managed DB (Supabase/PlanetScale/MongoDB Atlas) + a managed auth provider (Clerk/Auth.js) — don't build custom infra for an unvalidated idea.
• Deploy to Vercel — zero-config CI/CD, don't hand-roll deployment pipelines pre-validation.
• Payments: a Stripe Payment Link or Checkout (see payment-gateway-integration skill) rather than a full billing system, until there's proven demand.
• Skip: microservices, custom design systems, i18n, complex RBAC, extensive test coverage — all premature before product-market fit, and expensive to build for something that may pivot.