What problem does it solve? Building and operating multi-task data workflows on Databricks requires knowing the correct job configuration syntax across three different interfaces (Asset Bundles, Python SDK, CLI), plus triggers, compute, retries, and monitoring. This Skill provides the complete reference and working patterns so jobs are configured correctly before implementation begins. ## Core Features & Use Cases - Job Scaffolding and Deployment: Scaffold new job projects with databricks bundle init, then validate, deploy, and run via Declarative Automation Bundles across dev, staging, and prod targets. - Task and Trigger Configuration: Configure all task types (notebook, Python wheel, SQL, dbt, pipeline, JAR, run_job, for_each) and triggers (cron, periodic, file arrival, table update, continuous) with DAG dependencies and run_if conditions. - Monitoring and Reliability: Set up email and webhook notifications, health rules, timeouts, retries, and run queue settings for production workloads. - Use Case: Build a daily ETL pipeline where three extract notebooks run in parallel on a shared job cluster, a transform task waits for all extracts, and failures trigger PagerDuty webhooks with automatic retries. ## Quick Start Ask the agent to create a Databricks job that runs an extract notebook followed by a transform notebook on a daily cron schedule using Asset Bundles.
