Deploy a schema safely

Deployment executes generated SQL against a database. Unlike validation and rendering, it needs an installed and registered ADO.NET provider.

Review a dry run

Set DUBURL_DATABASE_URL in your environment to the target connection URL. Select it by environment-variable name so credentials do not appear in process arguments:

duburl schema deploy schema.yaml --connection-env DUBURL_DATABASE_URL --dry-run

--dry-run renders the SQL without opening a connection. Check the target dialect, identifiers, and requested changes before continuing.

Apply the schema

duburl schema deploy schema.yaml --connection-env DUBURL_DATABASE_URL

Dropping existing objects requires --drop-existing; non-interactive execution additionally requires --yes. Use destructive options only when replacing those objects is intended.

Embed or automate deployment

Use the schema application API from .NET.
Use the automation guide from CI or a script.
Use the JSON-RPC worker from a persistent client. Worker live deployment is disabled by default and has its own opt-in policy.