verification-before-completion

To prevent premature completion claims in PR workflows, require fresh verification evidence before asserting success. Map each claim to a concrete command, run it fully, inspect the exit status and failure counts, and only communicate conclusions after confirming the output.

Evidence-based reporting for release discipline is a gate process that blocks status updates or task completion messages unless backed by fresh full command output. It prevents rationalized claims by mapping statements of success to concrete verification commands and inspecting their results.

Verify test validation and CI readiness by running the full relevant command locally, inspecting the output for failure counts, and checking the exit code. Only report a passing status if the fresh full output confirms zero failures and a successful exit status.

Yes, the verification gate works for linting and build validation. It applies consistent evidence patterns across testing, linting, builds, regression cycles, and requirements checklists, requiring full command execution and exit status inspection before any completion message is sent.

Stop anti-rationalization excuses like 'should work now' by enforcing a verification gate that blocks trust in agent reports without independent verification. Require identifying the specific proving command, running it fully, and checking exit codes before communicating conclusions.

Relying on confidence instead of fresh verification output leads to dishonest or unreliable claims. The limitation is that without running a concrete full command and inspecting its exit code and failure counts, any statement of fixing or passing remains unverified and prone to rationalization.