data-quality-management

Data quality is measured across six dimensions: completeness, validity, consistency, uniqueness, timeliness, and lineage. This Skill profiles critical data elements, scores their impact on decisions, and produces evidence-backed findings with confidence levels.

Trace the defect from the reporting layer back through transformations to the source system, then remediate at the source rather than patching downstream. The Skill's framework profiles data, scores impact, traces root cause, and defines monitoring controls.

Run data quality diagnosis before committing resources to any material decision, when a KPI or event suggests data constrains an objective, or when the founder requests a review. Missing facts that could reverse a decision should stop the process.

It cannot make legal, tax, or regulated determinations requiring licensed specialists, and it must not invent benchmarks, costs, or probabilities. Actions involving money movement, external communication, or production changes require human approval.

Track critical-data quality scores, defect recurrence, and time to remediation, each with baseline, target, actual, trend, confidence, and owner. Also monitor recommendation calibration comparing expected versus actual results.