Ship-It Kit for agent-built apps

Your agent writes the code. This gets it shipped.

Version-checked agent rules, 25 build prompts, 8 rescue prompts and a 20-point security audit for Cursor, Claude Code, Copilot and Codex. One download for every project you build.

US$17 US$22

Buy now

Prices in US dollars. Your bank converts at checkout.Made in New Zealand by Vibemast.

See what is inside
  • Instant download
  • One-time payment
  • Not working? Fixed or refunded within 14 days

Works with

  • Cursor
  • Claude Code
  • GitHub Copilot
  • Codex
  • Gemini CLI
  • Windsurf
Vibemast Ship-It Kit cover: agent rules and prompts for AI apps
Ship-It Kit

US$17 US$22

What is in the kit

Everything your agent needs, in one download

Plain text files you drop into a project. No account, no subscription, no app to install.

  • 01

    Agent rules for 4 stacks

    Next.js + Supabase, Vite + React + TypeScript, FastAPI and Phaser. Always-on rules that sit in your project and shape every session.

  • 02

    One generator script

    Writes AGENTS.md, CLAUDE.md, Cursor rules and Copilot instructions from one source, so every tool follows the same rules. Copy-and-paste versions included.

  • 03

    25 build prompts

    Plan, scaffold, auth, database, payments, deploy. Each one makes the agent plan first, change little and prove it works.

  • 04

    8 rescue prompts

    For when the agent loops, breaks things, or leaves a mess you cannot read. Pick the one that matches your symptom.

  • 05

    20 security audit prompts

    Plus a one-page printable pre-launch checklist: secrets, access rules, auth, injection, payments, uploads, rate limits and AI-feature risks.

  • 06

    3 agent roles

    Planner, builder and reviewer, so the same chat does not plan, build and grade its own work.

  • 07

    Quickstart and licence

    A README that gets you set up in about three minutes, and a plain-English licence for personal and commercial use.

Checked against current versions on 27 Oct 2026: Next.js 16.4, React 19, Vite 8.3, TypeScript 7, Tailwind 4, FastAPI 0.162 with Pydantic 2, Phaser 4.2.1 and current Supabase keys.

Stuck or shipped

The last mile is where agent-built apps stall

Stuck

  • The agent rewrites 40 files to fix one bug.
  • It uses an API that does not exist in your version.
  • You paste the same fix request for the fourth time.
  • Keys, tokens and open database rules slip into production.
  • The same chat plans, builds and grades its own work.
  • It works on your machine and fails in production.

With the kit

  • Rules keep changes small and ask the agent to prove each one runs.
  • It is told to check your installed version before it writes code.
  • A rescue prompt changes the shape of the request and breaks the loop.
  • A 20-point audit looks for them before real users do.
  • Planner, builder and reviewer roles keep those jobs apart.
  • A local-versus-production prompt finds the difference, and a deploy prompt walks you through launch.

A look inside

One of the 8 rescue prompts, free to copy

Symptom: the agent has made three or more attempts at the same problem and keeps apologizing. When a fix has failed twice, change the shape of the request.

Why it works. It forces facts over guesses and gets the agent out of the rut.

Rescue prompt 1: Break the loop 
Stop. Do not change any code.
You have tried to fix [PROBLEM] several times and it is still failing. Write a short report:
1. The exact error or wrong behavior, copied word for word.
2. Every fix you have tried so far and why each failed.
3. What you now know FOR CERTAIN (verified by running something) versus what you are ASSUMING.
4. Three different hypotheses, including at least one that contradicts your earlier assumptions.
5. For each hypothesis, a read-only check or tiny experiment that would confirm or kill it.
Then run the cheapest experiment and tell me the result. Make no fix until one hypothesis is confirmed.

Who it is for

Built for people who ship with AI agents

It is for you if

  • You build apps with Cursor, Claude Code, Copilot, Codex, Gemini CLI or Windsurf.
  • You are a solo builder, indie hacker or designer who wants fewer surprises: fewer invented APIs, fewer giant rewrites, fewer leaked keys.
  • You use Next.js with Supabase, Vite and React, FastAPI, or you are making a browser game with Phaser.
  • You are comfortable running an AI coding agent and pasting a command into a terminal.

It is not for

  • Complete beginners who have never opened a code editor.
  • Teams that need a formal security assessment. An AI audit is a strong first pass, not a guarantee.
  • Anyone looking for a SaaS tool or app. This is a download of rules, prompts and checks.

If it does not work for you, tell us within 14 days

It is an instant download, so we cannot take it back. If the kit does not work for you or is not as described, contact us within 14 days of buying and we will make it right.

Read the refund policy

Quick facts

The Ship-It Kit at a glance

Price
US$17 one-time (regular price US$22)
Format
One ZIP of plain text files: Markdown, a short Node.js script and a one-page PDF checklist. No account and no subscription.
Delivery
Instant download after payment, from the order confirmation page and email.
Contents
Agent rules for 4 stacks, 1 generator script, 25 build prompts, 8 rescue prompts, 20 security audit prompts, 3 agent roles, README and licence.
Works with
Cursor, Claude Code, GitHub Copilot, OpenAI Codex, Gemini CLI, Windsurf and any tool that reads AGENTS.md.
Checked against
Next.js 16.4, React 19, Vite 8.3, TypeScript 7, Tailwind 4, FastAPI 0.162 with Pydantic 2, Phaser 4.2.1 and current Supabase keys, checked on 27 Oct 2026.
Licence
Use it in as many of your own and your clients' projects as you like. No resale or redistribution of the kit itself.
Updates
Refreshed when a major framework release changes the rules. The download link in your order email gets the new file.
Refunds
Contact us within 14 days if it does not work for you or is not as described.
Limits
An AI audit catches common mistakes. It does not replace a professional security review.

Questions

Frequently asked questions

What do I actually get?

A single ZIP file of plain text files (Markdown, a short Node.js script and a PDF). No account and no subscription.

Which coding tools does it work with?

Cursor, Claude Code, GitHub Copilot, Codex, Gemini CLI, Windsurf and anything that reads AGENTS.md. The script writes the right file for each.

Do I need to know how to code?

You need to be comfortable running an AI coding agent and pasting a command into a terminal. The README walks you through setup in about three minutes.

Which versions are the rules written for?

Next.js 16.4, React 19, Vite 8.3, TypeScript 7, Tailwind 4, FastAPI 0.162 with Pydantic 2, Phaser 4.2.1 and current Supabase keys, checked against those releases on 27 Oct 2026. Each rules file also tells the agent to check your installed version first.

Is this just a pile of prompts?

No. The core of the kit is the always-on rules that sit in your project and shape every session. The prompts, rescue guides and audit sit on top of them.

Can I use it for client work and commercial apps?

Yes. Use it in as many of your own and your clients' projects as you like. You may not resell or redistribute the kit itself.

Will it get updates?

Yes. When a major framework release changes what the rules should say, we refresh the kit and replace the file behind your download link, so the link in your order email gives you the new version. If that link stops working, email hello@vibemast.com and we will send a new one.

Is there a free sample?

Yes. The free mini pack includes the core rules, one stack and five prompts. You can get it with the form on this page.

Can I get a refund?

Because this is an instant download, contact us within 14 days if it does not work for you or is not as described, and we will make it right.

Can an AI audit replace a security review?

No. It catches the common mistakes that sink agent-built apps. For payments at scale or health, children's or other regulated data, get a professional review.

Before you buy

Questions and background reading

Also from Vibemast

Run a business, not just code?

The Business Brain Kit makes ChatGPT, Claude or Gemini know your business: what you sell, your prices, your tone and what you never promise. No coding needed. Start with the free setup prompt.

Get the free setup promptSee the Business Brain Kit

Hoist it. Ship it.

Get the rules, prompts and audit that keep your agent on course, then get your app in front of real users.

US$17 US$22

Buy now

Prices in US dollars. Your bank converts at checkout.Made in New Zealand by Vibemast.

Instant download. One-time payment. Not working? Fixed or refunded within 14 days.