How to use react-starter-kit: setup, run and deploy

What you need before you start

The README lists these prerequisites. The kit is built around Bun, so Node.js and npm are not part of the setup.

• Bun v1.4+ (the README says Bun replaces Node.js and npm)
• VS Code with the project's recommended extensions (recommended, not required)
• React Developer Tools browser extension (recommended)
• A Cloudflare account, needed for deployment

Create your project and install dependencies

Generate a new repository from the template on GitHub, then clone it locally. Replace the placeholder URL with your own repository's URL.

Next, install dependencies from inside the project folder.

Configure environment variables

The project follows Vite's env conventions. The committed .env file holds shared defaults and placeholders only, with no real secrets. Your real credentials go in .env.local, which is git-ignored. Values in .env.local override those in .env.

Copy the defaults into a local file and then replace the placeholder values with real ones. Also check apps/api/wrangler.jsonc for Worker configuration and bindings.

• .env is committed and holds placeholders
• Put real credentials in .env.local, which is git-ignored
• Values in .env.local override .env

Start the apps in development mode

Running bun dev starts all three apps (web, api and app). You can also start them one at a time. The README lists the local URLs for each app.

• React app: http://localhost:5173
• Marketing site: http://localhost:4321
• API server: http://localhost:8787

Set up the database

Make sure DATABASE_URL is set in .env.local. The README uses Drizzle ORM with Neon PostgreSQL. For quick local setup, push the schema directly. Staging and production should apply migrations instead, so use push only for local prototyping.

• bun db:push pushes the schema directly (local prototyping only)
• bun db:seed loads sample data (optional)
• bun db:studio opens a database GUI
• Staging and production use bun db:migrate:staging and bun db:migrate:production

Deploy to Cloudflare Workers

Each app deploys independently to Cloudflare Workers. The README gives a short production outline; the full steps are in the deployment guide at reactstarter.com. First, Terraform provisions the Hyperdrive configurations the API worker uses. Paste the two IDs it prints into the hyperdrive block of apps/api/wrangler.jsonc.

Then set production secrets on the API worker. BETTER_AUTH_SECRET and RESEND_API_KEY are required. Google sign-in and Stripe billing are optional. Set RESEND_EMAIL_FROM to an address on a domain you have verified with Resend, because the default onboarding@resend.dev only delivers to your own inbox.

• Provision infrastructure with bun infra:production apply
• Add the printed IDs to the hyperdrive block in apps/api/wrangler.jsonc
• Set required secrets: BETTER_AUTH_SECRET and RESEND_API_KEY
• Optional Google secrets: GOOGLE_CLIENT_ID and GOOGLE_CLIENT_SECRET
• Optional Stripe secrets: STRIPE_SECRET_KEY, STRIPE_WEBHOOK_SECRET, STRIPE_STARTER_PRICE_ID, STRIPE_PRO_PRICE_ID (set all four or none)
• Migrate production with bun db:migrate:production, which reads .env.production.local