djouallah

How to build, configure, and run dbt projects on the duckrun adapter — dbt models executed in DuckDB and materialized as Delta Lake tables via delta-rs, locally or on S3/GCS/ADLS/OneLake (Microsoft Fabric) — plus the SQL-first duckrun.connect() notebook API over the same tables. Use this skill whenever a data engineer is setting up a duckrun profile, writing models or sources for it, choosing an incremental strategy (merge, insert, append, delete+insert, microbatch), pointing dbt at OneLake/a Fabric Lakehouse, using duckrun.connect() in a notebook, debugging a duckrun run, or asking "dbt + DuckDB + Delta" questions in general. Consult it BEFORE writing profiles.yml or any incremental model config — several defaults differ from other dbt adapters.