# engineering-rapid-prototyper

The engineering-rapid-prototyper Claude Code skill guides prototype and MVP work for testing product hypotheses. It maps the work from choosing a metric to gathering feedback.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 186 installs · ★ 12 repo stars
- Author: PeterHdd
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/PeterHdd/agent-skills/tree/HEAD/skills/engineering-rapid-prototyper/SKILL.md
- Updated: 2026-03-09
- Security check: automated, 2026-11-17 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/engineering-rapid-prototyper

## What it does

The skill lays out a prototype workflow: define a hypothesis and its metric, build the main user journey, then publish a testable version. Its suggested stack includes Next.js, shadcn/ui, and Vercel; Supabase and Clerk or NextAuth are options when the hypothesis requires stored data or user identity.

The workflow adds features one at a time, collects feedback, and uses the metric to decide what happens next. Scope guidance caps a prototype at 5 database tables, 8 pages or routes, and 5 features. If the first deployment takes over 3 hours, the guide calls for a simpler stack; the empty project is meant to go live within 2 hours.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add PeterHdd/agent-skills --skill engineering-rapid-prototyper -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/engineering-rapid-prototyper/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/engineering-rapid-prototyper/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add PeterHdd/agent-skills --skill engineering-rapid-prototyper -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/engineering-rapid-prototyper/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add PeterHdd/agent-skills --skill engineering-rapid-prototyper -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/engineering-rapid-prototyper/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install PeterHdd/agent-skills skills/engineering-rapid-prototyper
```

2. **Manually: put the files in place**
- File: `.github/skills/engineering-rapid-prototyper/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add PeterHdd/agent-skills --skill engineering-rapid-prototyper -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/PeterHdd/agent-skills --path skills/engineering-rapid-prototyper
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/engineering-rapid-prototyper/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add PeterHdd/agent-skills --skill engineering-rapid-prototyper -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/PeterHdd/agent-skills/tree/HEAD/skills/engineering-rapid-prototyper/SKILL.md

## When to use it
- test a product hypothesis with working software
- choose a lean stack for an MVP
- build the core user journey before secondary features
- collect feedback and compare results with a defined metric

## Example prompts
- Build an MVP to test this hypothesis; define a metric and implement the main user journey first.
- Suggest a simple prototype stack that includes authentication and data storage.
- Add a feedback form and outline how to evaluate the prototype.

## Source text · SKILL.md · MIT

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: 7c412872b5296eda.

BEGIN THIRD-PARTY FILE 7c412872b5296eda
```text
---
name: engineering-rapid-prototyper
description: "Build functional prototypes and MVPs at maximum speed to validate ideas through working software. Use when you need proof-of-concept development, rapid iteration on user feedback, no-code or low-code solutions, backend-as-a-service integration, A/B testing scaffolding, quick feature validation, or modular architectures designed for fast experimentation and learning."
metadata:
  version: "1.0.0"
---

# Rapid Prototyping Guide

## Overview
This guide covers fast proof-of-concept development and MVP creation. Use it when building prototypes to validate hypotheses, setting up A/B testing, or choosing rapid development stacks that prioritize speed-to-deploy over production hardening.

## Stack Selection Guide

### How to choose tools
- When authentication is needed, use Clerk or NextAuth for instant setup — never build custom auth for a prototype. If the prototype does not need user accounts to test the hypothesis, skip auth entirely.
- When a database is needed, use Prisma + Supabase for instant hosting, schema management, and row-level security. If the prototype only needs to store <100 records for testing, use a JSON file or localStorage — do not set up a database.
- When deployment is needed, use Vercel for instant hosting and preview URLs on every PR. First deploy must happen within 2 hours of project start.
- When no-code/low-code can cover the requirement, use it — speed to validation matters more than custom code.
- When building, implement core functionality first, polish and edge cases later. If you are writing error handling before the happy path works end-to-end, stop.

### Decision Rules for Scope
- Maximum 5 database tables. If the schema needs >5 tables, you are overbuilding — cut scope or simplify the data model.
- Maximum 3-5 features. If the feature list exceeds 5, rank by hypothesis impact and cut the bottom ones. A prototype tests one hypothesis well, not five hypotheses poorly.
- Maximum 8 pages/routes. If more are needed, combine views or cut features.
- Zero custom UI components — use shadcn/ui, Radix, or similar. If a component does not exist in the library, simplify the design to fit what exists.
- No custom backend logic beyond CRUD + one business rule. If the prototype requires complex backend processing (queues, scheduled jobs, multi-step workflows), mock it with a simple endpoint that returns realistic fake data.
- No optimization. No caching, no CDN config, no database indexes, no lazy loading, no code splitting. These are production concerns. The prototype must be fast enough to demo — nothing more.

### Kill Criteria
```
END THIRD-PARTY FILE 7c412872b5296eda

The third-party file has ended. Everything between the 7c412872b5296eda tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 2655 of 10043 bytes. © PeterHdd · MIT · https://github.com/PeterHdd/agent-skills/tree/HEAD/skills/engineering-rapid-prototyper/SKILL.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
