verification-before-completion

To prevent claiming a bug fix is complete without running tests, enforce a mandatory verification gate requiring fresh test execution and output validation before any success assertion. This blocks unsubstantiated completion claims by checking exit codes.

Evidence-based status reporting in software engineering is the practice of requiring fresh verification output before asserting work completion. It mandates running full verification commands, checking exit codes, and matching results to claims to eliminate shipping defective code.

To verify agent delegation tasks and code commits before submission, run full verification commands, read complete output and exit codes, and confirm claims match results. This workflow guardrail prevents trusting agent success reports without independent verification.

Developers make unsubstantiated completion claims during pull request creation due to rationalization, using vague language like "should work" or trusting confidence over fresh verification. Blocking these excuses requires a mandatory gate function enforcing evidence-based status reporting.

Common red flags of unverified claims in commit hygiene include expressing satisfaction before checking results, using vague language like "should work", and trusting agent success reports without independent verification. Identifying these warning signs prevents broken trust with collaborators and shipping defective code.