Commands

What is a run?

A run is a single execution of your pipeline. It can include all assets, a single asset, or a subset (e.g., one asset plus its downstream dependencies).

bruin run

The most basic invocation runs the entire pipeline with default settings:

bruin run pipeline.yml

Common flags

Flag	Purpose
`--start-date`	Set the run start date
`--end-date`	Set the run end date
`--full-refresh`	Drop and recreate tables instead of incremental insert
`--environment` / `-e`	Choose which environment to use (e.g., `dev`, `prod`)
`--var`	Override custom variables at runtime
`--downstream`	Also run all downstream assets

Example with flags

bruin run \
  --start-date 2024-01-01 \
  --end-date 2024-02-01 \
  --full-refresh \
  --var taxi_types='["yellow","green"]' \
  -e default \
  path/to/pipeline.yml

Without --full-refresh, an append-materialized asset will only insert new rows. With it, Bruin drops the table and recreates it from the query results. This is useful for resetting tables or first-time runs.

Running from the VS Code extension

The Bruin panel lets you configure all the same options visually - set dates, toggle full refresh, add variable overrides, and choose to run a single asset, an asset with downstream, or the full pipeline.

bruin validate

Validates your pipeline before running:

bruin validate path/to/pipeline.yml

This checks:

Asset definitions and types are correct
Dependencies exist and there are no circular references
SQL syntax is valid (dry-run on BigQuery and Snowflake)
Connections and environment configuration are correct

Always validate before deploying to catch issues early.

Key points

bruin run executes your pipeline with configurable dates, environments, and variable overrides
--full-refresh recreates tables from scratch instead of appending
bruin validate catches configuration, syntax, and dependency errors before runtime
Both commands work from the CLI and the VS Code extension panel