What could a first pilot reveal about handling work returned by another team?

A project coordinator transfers engagement `ENG-2608` to the delivery team. The receiving lead returns it because the acceptance criteria lack a required sign-off. The scope and resource allocation remain valid, and the original delivery request remains open.

The pilot gives the evaluated agent the returned transfer, the missing-sign-off reason and the completed preparation records. Its task is to obtain the required approval through the permitted tool and resubmit the same transfer lineage. Creating a new engagement would detach the repair from the existing allocation and leave duplicate work for the coordinator.

A buyer can ask for this generated case alongside a variant in which the rejection invalidates the scope itself. The pair tests whether the agent distinguishes a repairable omission from a change requiring broader reconsideration.

Before expanding the pilot, inspect whether the tools expose enough information to make that distinction and whether the evaluator recognises both outcomes. The useful first result is a clear account of why the agent repaired, reopened or replaced work, with those actions tied to the rejection's actual effect.