title: Quickstart description: Create and run your first hatk project in under two minutes.#

Prerequisites#

• Node.js 22+ — check with node --version
• Docker — needed to run the local PDS (Personal Data Server) during development

Create a new project#

npx hatk new my-app

This scaffolds a full-stack project with a SvelteKit frontend, example feed, seed data, and everything wired together. The generated project includes:

my-app/
├── app/               # SvelteKit frontend (routes, components, styles)
├── server/            # Backend logic (feeds, XRPC handlers, hooks)
├── lexicons/          # AT Protocol schemas for your data types
├── seeds/             # Test fixture data for local development
├── hatk.config.ts     # Project configuration
├── hatk.generated.ts  # Auto-generated types from your lexicons
├── vite.config.ts     # Vite + SvelteKit config
└── docker-compose.yml # Local PDS for development

Start the dev server#

cd my-app
npm run dev

This does three things automatically:

1. Starts a local PDS via Docker (your own mini AT Protocol server)
2. Runs seed data to create test accounts and records
3. Starts the hatk backend and SvelteKit dev server with hot reload

See it running#

Open http://127.0.0.1:3000 in your browser. You should see the starter app running with seeded data. Both the frontend and API are served on the same port.

Try hitting an endpoint directly:

curl http://127.0.0.1:3000/xrpc/dev.hatk.describeCollections

This returns the data collections your server is indexing, derived from your lexicon schemas.

Next steps#

• Project Structure -- understand each file and directory
• Configuration -- customize hatk.config.ts
• Feeds -- build custom feed algorithms
• Auth -- add login and authenticated actions