data-incident-response

Root-cause analysis for a dbt pipeline failure is executed by running test-results.js and source-freshness.js to surface failing models and stale sources, then mapping the incident timeline to pinpoint the exact data quality breakdown.

Blast-radius mapping for data incidents identifies all downstream metrics and dashboards impacted by a data quality anomaly. It uses manifest-lineage.js to trace dependencies and prevent affected data from reaching decision-makers.

Triage dashboard drift and metric anomalies by phasing the investigation: identify the alert source, run manifest-coverage.js to check test gaps, isolate the failing model, and assess the scope of the data incident before remediation.

Write a postmortem test after a data quality incident by defining validation checks that capture the specific failure mode, then add it to your dbt suite to prevent regression. The workflow outputs a remediation plan and postmortem tests.

Yes, a data-stack-context is required to use this incident response workflow. It integrates with CLI tools like manifest-lineage.js and source-freshness.js to provide the necessary metadata for triaging and resolving data incidents.