verification-before-completion

Verification-before-completion requires you to identify and execute the proving command—such as running your test suite or build step—inspect the complete output and exit status, and confirm the results substantiate your claim before asserting success. This prevents incomplete work from advancing and eliminates premature declarations.

Enforce verification-before-completion across your workflow: run fresh verification commands for every success claim, review full output rather than assuming results, and back all completion reports with concrete evidence. This practice applies universally to feature implementations, bug fixes, and test runs.

Tentative language signals unverified assumptions. Verification-before-completion flags these red flags and requires actual proof—run the test, review the output, and present passing results as concrete evidence instead of speculation about what might work.

Stop premature declarations by requiring fresh, verifiable evidence before any completion claim. Execute the relevant test command, inspect the full output and exit status, and confirm it substantiates your work. This eliminates incomplete or faulty work passed on to downstream teams.

Yes. Verification-before-completion applies universally to software development, CI/CD pipelines, QA testing, data analysis, and general task execution. The core principle—prove claims with fresh evidence—has no exceptions and works across any success communication or completion report.

Acceptable proof is the complete output and exit status of the proving command you identified for the task—such as passing unit tests, successful build logs, or validation script results. The output itself must substantiate your specific completion claim; tentative satisfaction or partial inspection does not qualify.