Why train an agent to wait when something is already late?

**Why should waiting count as successful agent behaviour?**

Because the purpose of follow-up is to move an obligation towards resolution under a policy. Additional contact can obstruct that purpose when another authorised action is already in progress.

**What does that look like in a payment workflow?**

A utilities account has a late balance and an accepted promise to pay. The policy states that routine reminders remain suspended until the promise review point. The balance is still overdue, but the commitment changes which actions are appropriate. Sending another reminder merely because the balance remains unpaid disregards the active arrangement.

**What should the agent do instead?**

It should verify that the commitment applies to the outstanding balance, record why contact is deferred and preserve the review trigger. Waiting without a future trigger would leave the obligation unattended.

**How could generated material support this?**

MirrorFirm could generate tasks where the correct outcome is a documented deferral with a scheduled reassessment. Training can then distinguish purposeful waiting from omission, and distinguish a valid commitment from an informal note that does not authorise a pause.