verification-before-completion

To prevent false completion claims before opening a pull request, require fresh verification evidence by running the full command and reviewing complete output. This process enforces exit-code and failure counting to validate completion status.

Evidence-based verification for CI workflows is a gatekeeping process that requires full-command execution and complete output review before claiming done. It validates bug fixes, linters, and builds by checking exit codes and failure counts.

Common red flags when checking if tests pass include relying on should or probably statements, performing partial checks, or trusting an agent's success report without independent verification. Full execution and output review are required for valid completion evidence.

To verify a bug fix is ready for a PR, run the full verification command fresh and review the entire output. You must confirm exit codes and failure counts provide evidence-backed confirmation that satisfies your intended completion claim.

Good completion evidence for linters and builds requires running the full command fresh and reviewing complete output. It involves checking exit codes, counting failures, and confirming the results satisfy the specific readiness claim before stating success.