verify

Test result verification requires running specific test suites and confirming their output and exit codes show zero failures before any completion claim can be made. This prevents committing code that appears fixed but hasn't been re-verified.

An evidence before assertions policy requires fresh, executed command output backing any claims of passing tests, clean linters, or successful builds. It combats dishonesty and inefficiency by preventing premature or unverified work completion claims during code review.

To prevent false positives when validating build success and linter status, you must run the relevant commands and check their actual output and exit codes. This avoids assuming success just because code was changed.

Yes, automated task completion verification applies to agent task completion by enforcing a strict verification process. Agents must run specific commands and confirm fresh output before asserting any task is successfully finished.

A quality gate might pass prematurely if success is claimed without running the test suite to confirm zero failures. Enforcing mandatory verification of executed command output prevents committing code that hasn't been re-verified.

The best way to verify code changes before committing is to enforce a strict verification process that requires running specific commands and confirming their output and exit codes. This ensures all success claims are backed by fresh evidence.