verification-before-completion

To enforce fresh verification before claiming completion, apply a step-by-step gate that identifies the appropriate validation command, executes it, reads the full output, and validates the results before allowing any success claim.

Completion claims fail without recent evidence to eliminate premature or dishonest assertions that cause errors, requiring fresh test runs, linter checks, or build validations to be executed and confirmed prior to finishing.

The best way to verify bug fixes and prevent premature completion claims is implementing an evidence-based gate that runs the specific verification command, reads the output, and validates success before declaring the work finished.

Yes, you can use this verification gate for pull-request creation and regression checks, as it applies to any development task involving tests, linters, builds, or bug fixes where a success claim might be made.

To identify the right verification command before declaring work finished, follow the gate's guidance to determine the appropriate test, lint, or build command, execute it, and interpret the results to prevent shortcuts.

You should not use an evidence-based completion gate when a development task requires no validation commands, though it remains necessary for testing, building, linting, and pull-request tasks to ensure reliable completion claims.