verification-before-completion

You can enforce running tests before completion by applying a verification protocol that requires executing specific commands and presenting their output as evidence before any claims of success are made.

Evidence-based verification for quality assurance is a protocol requiring tangible proof, such as command output and exit codes, to support claims of completion or correctness before committing code or declaring a task done.

To stop premature declarations of done in software development, mandate the execution of verification commands like test suites and linters, then require the presentation of their successful output as proof of correctness before any assertions are made.

Yes, you can require command output as proof before committing code changes by enforcing a verification protocol that mandates running test suites and linters, then presenting their successful output to confirm the changes are complete and correct.

Yes, verification before completion applies to agent task completion scenarios by enforcing a strict protocol that requires the execution of specific commands and the presentation of their output as auditable proof before any claims of success are made.