Why not evaluate handovers using summary quality alone?

**Is a faithful summary enough to evaluate a handover?**

No. A summary can accurately describe previous work while omitting the distinction that determines what happens next. Readability and factual coverage are useful checks, but they do not establish that the receiving team can safely act.

**What distinction matters in practice?**

In healthcare administration, a scheduling team may receive a referral that has been administratively checked but has not received the required clinical review. A summary saying that the referral is complete can preserve much of the source content while incorrectly suggesting that booking may proceed. The key distinction is between complete paperwork and permission to schedule.

**How should evaluation teams use that difference?**

A generated task could ask an agent to prepare the transfer and then require a receiving agent to choose its next action from the transferred context. Evaluate both the message and the decision it enables.

This tests decision sufficiency: whether the recipient can identify permitted action, unresolved conditions and the appropriate escalation route. It also exposes summaries that sound decisive by removing uncertainty the recipient actually needs to retain.