Verification & Quality Assurance

Automated verification checks validate code correctness, security, best practices, performance, and documentation with real-time truth scores on a 0.0–1.0 scale. The Skill assigns reliability metrics to detect issues before they reach production, enabling CI/CD pipelines to reject code below your threshold.

Automatic rollback reverts changes that fail verification to the last known good state. When code scores below the default 0.95 truth threshold, the system instantly undoes the deployment, preventing faulty code from persisting in your repository.

Yes. The Skill exports truth scores and verification results for CI/CD monitoring and trend analysis. It works with git and node-based workflows to automate checks, log metrics, and trigger rollbacks as part of your deployment pipeline.

Truth scoring provides objective metrics on a 0.0–1.0 scale that measure how reliably code, agents, and task outputs function. The default 0.95 accuracy threshold flags outputs that need review or correction before deployment.

Verification checks validate security vulnerabilities, performance issues, documentation compliance, best practices, and code correctness. These automated checks identify gaps without manual review, ensuring consistent quality across all generated outputs.