A 311 resident-services desk has both draft agents and approved production agents. How should staff treat them?

A 311 resident-services desk should distinguish draft agents from approved production agents and only rely on approved ones for resident-facing work. Lifecycle hygiene depends on that separation: drafts are for building and testing; approved production agents are what residents and agents at the counter should trust. Treating the labels seriously prevents unverified behavior from reaching the public through 311 channels.

Distinguishing draft agents from approved production agents and sharing draft agents with residents with view-only access fails because drafts are unreviewed and must not be exposed to residents in any form, even view-only, since it can imply trust and leak unverified behavior. Distinguishing draft agents from approved production agents and enabling auto-publish of drafts after passing test chats fails because passing test chats do not constitute formal approval; a draft must complete the full review process before it is approved for resident-facing use. Distinguishing draft agents from approved production agents and tracking both by version labels for resident-facing work fails because labels alone do not authorize use; only approved production agents may handle resident-facing work while drafts must stay in building and testing.

Exam caveat: Domain 3 covers agent lifecycle concepts including draft versus approved/production states. On questions about how staff should treat agents, choose clear separation and production-only reliance for resident-facing scenarios.

Operational check: Inventory 311 agents by lifecycle state in admin surfaces, hide or restrict draft agents from resident-facing queues, train desk staff to launch only approved production agents, periodically reconcile labels after updates, and require re-approval when a draft fork is promoted so the distinction remains real rather than cosmetic.