verification-before-completion

Verification-before-completion enforces running fresh verification commands and capturing their exit codes and outputs as evidence before making any success claim. This prevents false positives in CI/CD pipelines and code reviews by requiring quantitative proof that tests, linting, and builds succeeded.

Implement a stepwise gate that identifies the proving command, executes it fresh, reads full outputs, verifies results match expectations, and only then claims completion. This three-tier model catches failures masked by cached or stale results.

Before declaring a feature ready, run mandated verification commands and capture their outputs as substantiation. Verification-before-completion ensures releases are backed by fresh, quantitative evidence from tests, linting checks, and build logs.

Yes. Verification gates enforce that reviewers see fresh test results and build outputs before approving changes. This workflow prevents merging code when verification commands have not run or their results are unread.

Cached results mask changes introduced after the original verification run. Fresh verification gates re-execute tests and linting every time before claiming success, ensuring regressions and new failures are detected at the point of completion.