reviewer-protocol

An enforced reviewer lockout automatically blocks the original author from producing the next version of a rejected artifact. The coordinator verifies author identity, refuses assignments to the original author, and selects a different agent for reassignment or escalation to maintain review independence.

When all eligible agents become locked out from revising a rejected artifact, the system handles the deadlock by escalating to the user. This prevents cycling or re-admitting locked authors, ensuring the multi-agent workflow halts gracefully rather than assigning revisions to unauthorized agents.

Yes, reviewer reassignment supports escalating rejected artifacts to a specialist agent. The coordinator verifies the selected agent is not the original author, allowing you to route rejected work to specialized agents for targeted revisions before escalating to the user if needed.

To set up a reviewer rejection lockout for QA testing workflows, instruct the coordinator to reject the artifact, lock out the original author, and assign a different agent to produce the next revision. The coordinator mechanically enforces this by verifying author identity before accepting any reassignment.

Multi-agent content moderation requires independent revision after a reviewer rejection to avoid defensive feedback loops. Enforcing mechanical lockouts ensures the original author cannot produce the next version, maintaining review independence and forcing the coordinator to select a different agent for the revision.