Data Quality Audit

Quick Start

Goal: find correctness issues quickly and return a small, actionable report.

When to use this skill

• bug reports that smell like data drift (missing rows, double counts, weird nulls)
• after migrations/backfills to prove invariants still hold
• when analytics numbers disagree between systems

Before you run checks

• database engine
• target tables + expected primary keys
• expected invariants (unique, not null, fk relationships, allowed ranges)
• performance constraints (large tables, peak hours)