verification-before-completion

Enforce verification before code completion by requiring fresh evidence from a clean state. You must identify the proof command, run it, read the output, and validate exit codes before making any success claim.

Evidence-based quality assurance for pull requests requires running tests, builds, and linting from a clean state. It blocks success statements until fresh verification output is explicitly validated, preventing unverified claims.

Stop unverified completion claims by applying an identify-run-read-verify gate. This protocol blocks any success statement until the full verification command is executed and its output confirms the result.

Yes, this verification protocol works for builds and linting checks. It covers typical software engineering workflows by requiring fresh test execution and output validation before claiming readiness for deployment or handoff.

The best way to verify test results is executing the full verification command from a clean state, reading the complete output including exit codes, and only then claiming success. This explicit gate prevents undetected issues.