verification-before-completion

Verification before completion requires running the relevant test or build command, capturing fresh evidence like output and exit codes, and checking the actual results before making any success claim. This workflow eliminates premature closure on unit tests, integration tests, linting, and builds.

The best way to verify build evidence is to execute the verification command, capture the full output and exit codes, and validate the actual results before claiming completion. This formal workflow prevents rationalization and ensures only validated work reaches deployment or pull requests.

Verification fails when completion is claimed without fresh evidence because it bypasses the required validation workflow. Guardrails against premature closure highlight red flags and prevent rationalization, ensuring that unit tests, integration tests, and builds are actually run and validated before any success claim.

To enforce QA testing evidence, state exactly what proves completion, run the verification command freshly, and report the actual results with evidence. This workflow applies to unit tests, integration tests, linting, and agent-driven tasks, preventing premature closure and unverified claims.

You need to run verification commands before claiming success for any task requiring validation, such as unit tests, integration tests, linting, or builds. The workflow demands fresh verification evidence, reading full output and exit codes, and confirming actual results before any completion claim.