Verification & Quality Assurance

You can automate quality verification in a CI/CD pipeline by running this Skill on pull requests. It scores reliability, checks best practices, blocks merges below threshold, and exports structured JSON reports for review.

Truth scoring assigns measurable reliability scores to files, tasks, and agent outputs. This mechanism replaces subjective code review with threshold-based metrics to quickly spot weak results and low-quality changes.

You should use automatic rollback when validation fails and a risky regression is detected. Git-backed recovery reverts the codebase to the last known good state, preventing bad edits from spreading through your pipeline.

Verification checks validate correctness, security, performance, and documentation quality across single files or whole directories. They apply threshold-based scoring and best-practice checks to ensure deployment readiness.

This Skill applies to verification, truth scoring, rollback, reporting, and dashboard workflows across files, directories, CI/CD pipelines, and live monitoring. It requires threshold-based scoring, security checks, structured JSON export, and Git-backed recovery.