verification-before-completion

To prevent premature code completion claims, enforce fresh verification evidence by identifying proving commands, running them fully, reading outputs, and verifying results before any success statements. This ensures agents produce auditable outputs before PRs or releases.

Evidence-based verification in CI-CD pipelines is the process of executing fresh test or build commands and reading their full outputs to confirm success. It enforces step-by-step verification to ensure accuracy and auditable traceability before releases.

To enforce step-by-step verification before a pull request, apply a gate function workflow: identify the proving command, run it fully, read the output, verify the result, and only then claim completion status to guard against premature conclusions.

Yes, you can use automated verification to improve quality assurance for builds by enforcing the iron law of always running fresh verification commands and reviewing full outputs before claiming success, ensuring accuracy and mitigating risk.

When verification tests fail during a build, the agent must handle the failure by recognizing it as a red flag, avoiding any success claims, and keeping an auditable log of the verification results for risk mitigation.

Yes, verification-before-completion applies to tests, builds, and code reviews by guiding agents to produce auditable outputs through fresh verification, ensuring completion is only claimed after confirming success.