Output Types

Every template produces a file, and the output type describes what kind of artifact that file is. It answers the question: once this is generated, what do I do with it? A table-creation script is deployed once and evolves through schema changes; a stored procedure is deployed and then called repeatedly by a scheduler or control framework; a view is deployed and then queried. Same metadata, very different operational lives.

Each template page in this reference states its output type, so you know up front what the generated result is and how it fits into a deployment.

The bundled sample templates use these output types:

• Data Definition Language (DDL): scripts that create or define data objects, such as tables or file structures. Run at deployment time to build or update the solution’s structure.
• Stored Procedure: executable database objects that encapsulate loading logic. Deployed once, then called repeatedly, typically by a control framework, to move data.
• View: stored queries that behave like tables. The backbone of the virtual data warehouse approach, where layers are delivered as views instead of physically loaded tables.

The set is open-ended. Because a template can generate any text, your own templates can target any output type that suits your stack: Markdown documentation, PySpark notebooks, dbt models, deployment manifests, and so on. The bundled Documentation template, for example, produces Markdown.