Shijie Rao 7f135e1314 Make Ultra reasoning fallback model-aware (#41206) ...

## What changed

- Add an optional `multi_agent_reasoning_effort` field to model metadata.
- When `Ultra` is selected, use a supported catalog override when present. Otherwise, prefer `max`, then the highest supported non-`ultra` effort, and finally `medium` when the model has no reasoning levels.
- Preserve existing request behavior for other reasoning efforts, including mapping `persistent` to `disabled`.

## Testing

- Cover configured, missing, unsupported, and empty-catalog fallbacks for both parent and spawned-agent requests.

GitOrigin-RevId: b7ef576f32e0ea95ef38bf95c25789ed16aae94e

2026-08-27 21:03:09 +00:00

..

clients.rs

Route Guardian inference through dedicated endpoints (#40892)

2026-08-26 14:36:48 +00:00

models_integration.rs

Make Ultra reasoning fallback model-aware (#41206)

2026-08-27 21:03:09 +00:00

realtime_websocket_e2e.rs

Reconnect WebRTC Realtime sideband transports (#39257)

2026-08-18 18:41:40 +00:00

sse_end_to_end.rs

Expose response usage metadata in completion events (#41087)

2026-08-27 07:50:05 +00:00