// the find
datacoves/dbt-coves
CLI tool for dbt users to simplify creation of staging models (yml and sql) files
A CLI that inspects your warehouse (Snowflake, Redshift, or BigQuery) and generates dbt source configs, staging models, and their property YAML files instead of you hand-writing boilerplate. It's built and maintained by Datacoves, a company selling a managed dbt/Airflow platform, and the CLI also bundles Airbyte/Fivetran config extract-and-load, Airflow DAG generation from YAML, and Snowflake blue-green deploys. Best fit for teams onboarding a lot of new sources into dbt who are tired of writing the same staging model shell by hand.
Test suite has real per-adapter fixtures (tests/generate_sources_cases/test_{bigquery,redshift,snowflake}_generate_sources) that diff generated SQL/YAML against expected output, not just mocked unit tests. Versioning is pinned to match dbt-core's major/minor from 1.4.0 onward, so you're not guessing compatibility. CI is mature — ruff, mypy, pre-commit, codecov, and automated tag/release/PyPI-publish workflows all wired up. Scope goes past model generation into genuinely adjacent chores: Airflow DAG generation from YAML and Airbyte/Fivetran config save-restore, useful if you're already gluing those tools to dbt.
The README opens with a full pitch for the commercial Datacoves platform before you get to what dbt-coves itself does — actual usage docs live one click away in docs/commands/. Warehouse support is hard-capped at Snowflake, Redshift, and BigQuery; Postgres, Databricks, and DuckDB users get an untested 'may work' shrug. 282 stars and 19 forks is thin for something that's tracked dbt-core since 1.0, so the less-central commands (blue-green, data-sync) have likely seen far less real-world battering than the source-generation path. Being a single vendor-backed project means the niche commands are one roadmap shift away from being deprioritized if they stop mattering to Datacoves' own platform.