skippr model
Model runs the data-engineer workflow for dbt planning, authoring, validation, review, and publish steps.
Usage
bash
skippr [--config <path>] [--log [level]] model --pipeline <name> [--no-resume]model derives warehouse settings from the selected pipeline's data_sink in skippr.yml and uses built-in defaults for catalog, dbt, vector, storage, and LLM settings. Extract and load are handled by skippr discover --pipeline <name> and skippr sync --pipeline <name>, not by the modeling workflow.
Flags
Command flags:
| Flag | Description |
|---|---|
--pipeline <name> | Required. Selects the configured pipeline to model; the warehouse comes from that pipeline's data_sink. |
--no-resume | Start a fresh modeling thread instead of resuming the latest project thread. Use this when the previous model run is stale, failed before useful state was persisted, or you deliberately want a clean attempt. |
Global flags:
| Flag | Description |
|---|---|
--config <path> | Read a config file other than ./skippr.yml. |
--log info | Structured logs to stdout |
--log debug | Debug-level logs |
--log trace | Maximum verbosity |
By default, skippr model --pipeline <name> resumes the latest thread for that pipeline when one exists. That makes retries idempotent and lets a run continue from persisted modeling state. Add --no-resume only when you want to discard that thread selection and start a new model run.
See also
skippr test— list and run dbt tests against the materialised project.skippr plan/skippr ask— planning and read-only Q&A without applying model changes.