New: the AI App Starter

New: the AI App Starter

Paste one line into Claude Code or Codex, and your AI coding agent builds a Next.js, Supabase and Vercel app with you, from idea to launch. It leads, in plain language:

• Safe by design. It works only on your computer, with a local database and stand-in payment and email services. It never holds keys, and pushing and deploying are blocked.
• Few questions, at the right time. It reuses your PRD and MVP scope and asks only what the next piece of work needs. The rest comes up when it matters.
• Organized from day one. One plan, one decisions file, separate local, staging and production environments, and fresh sessions that pick up where you left off.
• Honest about what's real. It checks current docs, says when a test used a stand-in, and never decides law, tax or pricing for you.

Tested so far: Claude Code desktop, through the first working pieces of a web shop. Codex support is in beta: supported, not tested yet. The hosted staging and production setup and the launch stage aren't tested yet either.

Get started: AI App Starter

Also changed

• The README is now organized around three parts: the planning prompts, the AI App Starter and agent configurations.
• The six planning prompts still work on their own in any AI chat. The starter runs the PRD, UX and MVP prompts inside your agent, or uses the documents you already made with them.
• Removed: the session continuation prompt, the v0.dev prompts and the permission guardrails config.