What problem does it solve? Building and maintaining dbt pipelines for BigQuery requires deep knowledge of SQL optimization, project configuration, and adapter quirks. This Skill guides an AI agent through creating, modifying, and troubleshooting dbt models with correct BigQuery-specific practices, reducing misconfigurations like missing profiles.yml files and unoptimized SQL. ## Core Features & Use Cases - dbt Project Scaffolding: Initializes new dbt projects with dbt_project.yml, profiles.yml, and proper folder structure under a dedicated directory. - SQL Optimization & Data Cleaning: Automatically applies BigQuery optimization rules (e.g., rewriting IN subqueries to EXISTS) and mandatory data cleaning protocols on every pipeline change. - Troubleshooting & Validation: Compiles projects with dbt compile, runs dbt test, and diagnoses remote orchestration failures (e.g., Cloud Composer DAG errors) via gcloud logging. - Use Case: Ask the agent to add a new staging model to an existing dbt project; it will inspect the DAG, fetch source table schemas with bq CLI, generate the model with a config block and unit tests, and validate everything with dbt compile. ## Quick Start Create a new dbt model in my project that aggregates daily revenue from the raw_orders BigQuery table and validate it compiles.
