High Model Failure Rate Alert

Monitor failure rates from dbt run history and test results. Alert when a model fails more than 2 times in a rolling 7-day window, or when test failure rate exceeds 10% over the same period.

Investigate logs in logs/dbt.log and target/run_results.json for error patterns. Check for upstream dependency failures, schema changes, or data quality issues. Use dbt test --store-failures to examine failing rows. Consider adding data quality checks or adjusting incremental logic.