Data Quality Issue Triage Planner

The problem

Vague data complaints like "the dashboard looks wrong" often lead to aimless debugging and wasted engineering hours. Teams struggle to bridge the gap between anecdotal reports and actionable, auditable technical requirements.

What it does

• Converts raw complaints and logs into structured issue definitions with unique item IDs for tracking.
• Builds an evidence ledger that maps symptoms to specific source timestamps, rows, or sections.
• Classifies data quality dimensions and maps upstream lineage without assuming system access.
• Generates diagnostic plans, repair tests, and monitoring proposals based on provided schema.
• Identifies evidence gaps, conflicting sources, and stale data instead of smoothing over missing info.

Why this beats prompting it yourself

Generic prompts often hallucinate missing context or provide vague "check the logs" advice. This skill enforces a strict evidence-based ledger that mandates citations for every claim, ensuring your triage is grounded in fact rather than inference.

Use cases

• Converting stakeholder bug reports into formal data engineering tickets.
• Triaging discrepancies between production databases and reporting warehouses.
• Conducting root cause analysis (RCA) for data pipeline failures or schema drift.
• Auditing third-party data feeds for consistency and reliability.

Known limitations

Cannot modify production data or perform automated backfills. Requires users to supply specific source materials or schema definitions to generate diagnostic plans.