quality-control

Automated accessibility checks validate production artifacts against WCAG 2.1 AA standards. The tool scans content and UI components, outputting structured JSON findings to enforce deterministic pass/fail quality gates before release.

SQLite audit logging persists quality control results through a quality_logger mechanism. It creates a traceable audit trail by storing structured validation findings locally in a SQLite database, ensuring long-term traceability of accessibility and brand checks.

Brand validation runs automatically against style bible markers to verify colors and typography. It compares artifact properties against predefined brand standards, generating deterministic pass/fail criteria that eliminate the need for manual visual review.

Deterministic quality validation enforces standard checks before release by running two-pass quality gates. It evaluates accessibility, brand compliance, and audit logging, returning structured JSON with clear pass/fail criteria for pre-checkpoint reviews.

YAML and Pillow are required dependencies for running automated brand validation. YAML handles configuration parsing for style bible markers, while Pillow processes visual artifact properties like colors and typography for compliance verification.

Deterministic quality checks are limited to validating accessibility, brand compliance, and audit logging for production artifacts. They rely on predefined style bible markers and WCAG 2.1 AA criteria, meaning subjective design evaluations remain out of scope.