Agent skill

setup-project

Initializes project with smart stack detection, skill recommendations, and standard config. Use when starting a new project.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/setup-project-djnsty23-claude-auto-dev

SKILL.md

Setup Project Workflow

On "set up" or "setup"

Step 1: Stack Detection

Scan for dependency files and detect the project stack:

Check for: package.json, pyproject.toml, Cargo.toml, go.mod

If package.json exists, read dependencies and devDependencies to detect:

Signal Detection
Framework next, react, vue, svelte, express, fastify, remix, astro
CSS tailwindcss, styled-components, @emotion/react
Database @supabase/supabase-js, prisma, mongoose, drizzle-orm
Auth next-auth, @supabase/auth-helpers, @auth/core, passport
Payments stripe, @stripe/stripe-js, @stripe/react-stripe-js
Testing vitest, jest, playwright, @playwright/test, cypress
Video remotion, @remotion/cli

Also check for config files:

  • vercel.json or .vercel/ -> Vercel deployment
  • netlify.toml -> Netlify deployment
  • Dockerfile -> Container deployment
  • supabase/ directory -> Supabase project
  • .github/workflows/ -> CI/CD pipelines

Step 2: Project Type Classification

Based on detected signals + user description, classify the project:

Type Signals
E-commerce stripe + product pages or store
SaaS auth + subscription/billing + dashboard
Marketing / Portfolio static pages, SEO focus, no auth
API express/fastify + no frontend framework
Full-stack app framework + DB + auth
Library / CLI no framework, bin field in package.json

If unclear, ask:

question: "What best describes this project?"
options:
  - { label: "E-commerce", description: "Online store with payments" }
  - { label: "SaaS", description: "Subscription app with auth" }
  - { label: "Marketing site", description: "Landing pages, SEO focus" }
  - { label: "Full-stack app", description: "Web app with database" }

Step 3: Generate CLAUDE.md

Create a project-level CLAUDE.md with:

markdown
# [Project Name]

## Overview
[One-line description from user or package.json]

## Stack
[Detected technologies]

## Key Commands
- `npm run dev` - Start development server
- `npm run build` - Production build
- `npm run test` - Run tests
[Add detected commands from package.json scripts]

## Key Directories
[Map out src/ structure if it exists]

## Patterns
[Stack-specific patterns based on detected technologies]

Step 4: Recommend Skills

Based on detected stack, list relevant auto-dev skills:

Always relevant: review, commit, fix

Conditional recommendations:

If Detected Recommend
@supabase/supabase-js supabase (db, rls, migrations)
stripe stripe (payments, webhooks, checkout)
next perf (Core Web Vitals), seo (meta tags, schema)
tailwindcss design (UI patterns, design tokens)
vercel.json deploy (Vercel deployment)
playwright/cypress test (unit + browser tests), browser (automation)
remotion remotion (video creation)
Marketing/E-commerce type seo (meta tags, schema markup)
Any auth detected security (pre-deploy scanning)

Output format:

Detected: Next.js + Supabase + Stripe + Tailwind

Recommended skills for this project:
- supabase — database, RLS policies, migrations
- stripe — payments, webhooks, checkout
- seo — meta tags, schema markup, Open Graph
- design — UI patterns, design tokens
- deploy — Vercel deployment
- security — pre-deploy scanning
- perf — Core Web Vitals

These load automatically when you use their trigger words.

Step 5: Environment Setup

Create .env.example based on detected services:

bash
# Generated based on detected stack

# Supabase (if @supabase/supabase-js detected)
NEXT_PUBLIC_SUPABASE_URL=
NEXT_PUBLIC_SUPABASE_ANON_KEY=

# Stripe (if stripe detected)
STRIPE_SECRET_KEY=
NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY=
STRIPE_WEBHOOK_SECRET=

# Auth (if next-auth detected)
NEXTAUTH_SECRET=
NEXTAUTH_URL=

# App
NEXT_PUBLIC_APP_URL=

Ensure .gitignore includes:

  • .env.local, .env
  • node_modules/
  • .next/ (for Next.js)
  • .claude/ (auto-dev artifacts)

Step 6: Create prd.json

Based on project type, create initial stories:

E-commerce starter:

  • Product listing page with schema markup
  • Shopping cart
  • Stripe Checkout integration
  • Order confirmation and webhooks
  • SEO: meta tags, Open Graph, sitemap

SaaS starter:

  • Auth flow (signup, login, logout)
  • Subscription billing with Stripe
  • Dashboard layout
  • User settings and profile
  • Protected API routes

Marketing site starter:

  • Landing page with hero section
  • SEO: meta tags, Open Graph, JSON-LD
  • Contact form
  • Responsive design
  • Sitemap and robots.txt

Step 7: Report

Project setup complete:
- CLAUDE.md: Created with [detected stack] context
- prd.json: [N] initial stories based on [project type]
- .env.example: Configured for [detected services]
- .gitignore: Updated

Recommended skills: [list]

Next: Say 'auto' to start working through tasks,
or 'brainstorm' to scan for more improvements.

Quick Setup

If user provides a description: "Set up this project - I want to build [description]"

  1. Run stack detection (Steps 1-2)
  2. Generate all files using smart defaults (Steps 3-6)
  3. Skip questions when the description provides enough context
  4. Show the report (Step 7)

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results