front: include running whisper-servers in /internal/engine-state

The supervisor polls /internal/engine-state (not /admin/api/status) to
populate each engine's loaded_models in the registry, which the front then
aggregates for the models page. That endpoint only read .models, so a
whisper-server running on a secondary engine still showed as not loaded.
Fold each running server (id + `audio:` alias) into its loaded_models.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

codai/api/app.py

@@ -247,6 +247,17 @@ async def internal_engine_state():
         loaded = list(multi_model_manager.models.keys())
     except Exception:
         loaded = []
+    # Whisper-server models run as their own subprocess (tracked in whisper_servers,
+    # not in .models). Surface each running server under both its id and `audio:`
+    # alias so the front's cross-engine aggregation shows it as loaded — including
+    # when it runs on a secondary engine.
+    try:
+        for _wid, _wsm in multi_model_manager.whisper_servers.items():
+            if _wsm.is_running():
+                loaded.append(_wid)
+                loaded.append(f"audio:{_wid}")
+    except Exception:
+        pass
     vram = None
     try:
         import torch