verify

You enforce evidence-based verification by running a deterministic workflow that identifies the proving command, executes it fully without truncation, reads the complete output, checks the exit code, counts failures, and confirms the result before making any completion claims.

The best way to verify code claims is to require fresh evidence in the same message as the claim by identifying a specific command, running it completely, reading the entire output, and validating that the exit code and failure count confirm the assertion.

An evidence-first verification process works by blocking completion or task closure until a full command is executed, its complete output is read, and the exit code alongside failure counts are checked to prove the claim before responding.

Yes, you can use this verification workflow for deployment and end-to-end testing scopes. It guides you through identifying, running, reading, and verifying full command outputs to ensure claims are backed by fresh evidence.

Completion claims fail quality assurance audits when they lack fresh evidence because there is no verifiable proof attached. Requiring full command execution and output validation in the same message ensures every claim is trustworthy and audited.

Relying on partial command output for process auditing is limited because truncated results hide failures and obscure exit codes. The verification workflow requires running the full command and reading the complete output to ensure accurate claim validation.