arc-verify

Verify task completion claims by selecting a proof command, running it to capture full output, and evaluating the results to determine if the claim is supported by fresh evidence. This prevents premature task closings.

Fresh command verification requires executing a chosen proof command and capturing its complete output to substantiate acceptance decisions for arc-tracked task claims. It eliminates ambiguity by backing every completion with new evidence.

Prevent premature arc task closings by using guardrails that ensure fresh proof output matches acceptance criteria before a completion claim is approved. The verification process reads complete command results to decide claim validity.

Yes, you can validate build and test outputs by selecting the relevant proof command, executing it completely within OpenCode, and evaluating the full output to determine if the verifiable results support the task claim.

Yes, arc verification operates without dependencies or external components, relying entirely on selecting a proof command, running it to produce full output, and evaluating the results to substantiate the acceptance decision.

The best way to evaluate proof command output is to run the selected command completely, capture the full results, and read the entire output to determine if the evidence matches the acceptance criteria for the arc-tracked claim.