Skip to content

Backend and data

Have a question? Ask ChatGPT Ask Claude

Every Regular project can have a managed backend — a database, user sign-in and a REST API — without you setting up a server. You find it in Dashboard → Dynamic Backend.

Usually you don’t do anything: when your app needs to store data, forms or user accounts, the AI defines the backend for you while it builds. Projects start without one and the backend is created when it is first needed (at the latest on the first publish).

You can also set it up yourself: Set up your data backend shows a starter schema; adjust it and press Create backend.

Your collections and their fields. Switch between Overview (a visual view) and JSON. For each field you see its type, whether it is required or unique, and its access rules:

  • Public — anyone can read it,
  • Owner — only the record’s owner,
  • Role — only users with a given role.

Change the schema and press Save schema.

Browse, add (New record), edit and delete records — in a form, or Edit as JSON.

The REST endpoints of your app, on your app’s own domain: list, view, create, update and delete for each collection, each marked Public, Signed-in required or Own records only. Calls authenticate as your app’s own users, not with a SKAEL login.

The people who sign up to your app — separate from the members of your SKAEL workspace. Turn Allow public sign-up on so anyone can register, or off to make your app invite-only. (It needs a collection that holds user accounts.)

Every backend includes end-user sign-in. The AI is set up to use it rather than writing its own login, so you get a tested sign-in without extra work.

Your sandbox and your live site have separate data. Publishing copies the schema, not the records — your live data is never overwritten by test data. To try real data in the sandbox, use Sync from prod (see The preview).

Export app includes every collection as CSV and JSON. See Import and export.