sql / dbt scaffolder

SQL → dbt Scaffolder

Paste raw SQL and get a starting dbt model — table references swapped for ref()/source() macros, plus a schema.yml stub with your columns listed.

Runs fully in your browser · nothing is sent anywhere

query.sql

Model name

Materialization

Detected tables

Paste SQL above to detect table references.

Waiting for input…0 chars

output

Paste a query and click Generate to scaffold a model.

models/my_model.sql

my_model.yml

—

How it works

Everything runs client-side — no network request, no upload, no server round-trip.
Uses the same tokenizer as our SQL Formatter/Explainer to find FROM/JOIN table references and the SELECT column list, while correctly ignoring CTE names defined in a WITH clause.
Each detected table gets its own control: wrap it in ref(), wrap it in source(), or leave it exactly as written. We can't know which tables are already dbt models vs. raw sources — that's your call.
The rest of your query (joins, filters, grouping) is preserved as-is; only the table identifiers are swapped.
Columns with an explicit AS alias are listed exactly; bare column references use their own name; anything more complex (an un-aliased expression, or SELECT *) gets a TODO flag since we can't always know the exact output column name.
Only the first statement is scaffolded if your input has more than one (separated by ;).