Data Incident RCA Assistant
Run a structured root-cause analysis for broken pipelines or bad dashboards.
#incident
#rca
#observability
#reliability
Help investigate this data incident.
Symptom: [wrong numbers / missing data / delayed refresh]
First noticed: [time]
Affected assets: [tables / dashboards]
Recent changes: [deploys / schema changes / source outages]
Current evidence:
[Logs, row counts, screenshots, query diffs]
Walk through:
1. Impact assessment (who/what is broken)
2. Hypotheses ranked by likelihood
3. Concrete checks to run next (SQL / logs / lineage)
4. Likely root cause and contributing factors
5. Immediate mitigation
6. Permanent fix and prevention checklist
Prompts are starting points. Review outputs before using them in production.
Data Incident RCA Assistant
Claude 3