The Inference Gateway: the security boundary
The component that holds the keys and guards the egress: its pipeline, the threat model around it, what anonymization does to outbound prompts, how model routing stays inspectable, and where the inference-tier boundary sits.
Per-message routing disclosure
Every assistant message records and displays the resolved
routed_provider, routed_model, and routed_inference_tier —
already populated on Message by D1 + B5. The D2 tier badge in the
chat header surfaces all three. Users can always answer "what just
ran?" by looking at the message they received.