data-quality-monitor-designer

What problem does it solve?

This Skill designs production data-quality monitoring so stale, missing, duplicated, invalid, inconsistent, or drifting data is detected before it harms reports, models, or downstream decisions.

Core Features & Use Cases

• Six-dimension monitoring design: Defines checks for freshness, completeness or volume, uniqueness, validity, consistency or referential integrity, and drift.
• Pipeline-stage placement: Places each check at the earliest viable stage, such as ingest, transform, or serving, to catch defects as soon as possible.
• Failure-action policy: Assigns each check a clear response: block, quarantine with provenance, or alert-and-pass, with no silent auto-fix.
• Operational metadata: Specifies severity, owner, per-dataset quality SLAs, incident routing, and residual coverage for gaps that remain.
• Use case: A warehouse feed starts losing rows on weekends and one tenant's data silently fails; this Skill designs the monitor spec, severity, owner routing, and quarantine policy that would surface it.

Quick Start

Design a data-quality monitoring spec for my critical production datasets, including checks, placement, severity, owner, failure action, SLAs, and quarantine policy.