fix: front-driven thermal pause indicator no longer erased by engine poll

On a global CPU cooldown the thermal supervisor already pauses ALL engines
(want_pause = gpu_hot or cpu_hot, applied per-engine), and the logs confirm
every engine gets a pause. But the Tasks page often showed only ONE engine
as cooling while the others looked like they were still running.

Cause: two threads write engine.cooling. The thermal loop sets it when it
pauses an engine; the health poll loop overwrites it every tick with the
engine's OWN self-reported cooldown (cooling=d.get("cooling")). An engine
only self-reports cooling while sitting in its own wait_until_safe loop (it
has an in-flight request). An engine the front paused while IDLE reports
cooling=None, so the poll loop cleared the front's pause indicator — the UI
then showed that engine as running mid-cooldown.

Fix: while the front holds an engine paused (engine.therm_paused), pass the
update_state sentinel (cooling=False, "don't touch") instead of the engine's
self-report, so the front's indicator survives. Once the front resumes the
engine, the engine's self-report flows through again as before.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LoSpEthysqmseCc6Geizty

codai/__init__.py

@@ -16,7 +16,7 @@
 
 # Canonical product version for CoderAI — single source of truth. Both the API
 # metadata and the admin web UI read from here.
-__version__ = "0.1.60"
+__version__ = "0.1.61"
 
 # Configure the CUDA caching allocator BEFORE torch is imported anywhere.
 # expandable_segments lets the allocator return freed pages to the driver even

codai/frontproxy/engine_supervisor.py

@@ -840,13 +840,26 @@ class EngineSupervisor:
                         d = r.json()
                         healthy = True
                         loaded = d.get("loaded_models") or []
+                        # Cooling indicator: the FRONT's thermal supervisor owns the
+                        # pause state (it's what pauses ALL engines on a global CPU
+                        # cooldown). The engine only self-reports cooling while it's
+                        # sitting in its OWN wait_until_safe loop (i.e. it has an
+                        # in-flight request). An engine the front paused while idle
+                        # reports cooling=None — so blindly taking the engine's report
+                        # here would ERASE the front's pause indicator and the Tasks
+                        # page would show that engine as running mid-cooldown. While
+                        # the front holds this engine paused, keep the front's
+                        # indicator (sentinel False = don't touch); only let the
+                        # engine's self-report through once the front isn't pausing it.
+                        _cooling = (False if getattr(engine, "therm_paused", False)
+                                    else d.get("cooling"))
                         self.registry.update_state(
                             engine.id, healthy=True,
                             loaded_models=loaded,
                             loaded_info=d.get("loaded_info") or [],
                             vram=d.get("vram"),
                             tasks=d.get("tasks") or [],
-                            cooling=d.get("cooling"),
+                            cooling=_cooling,
                         )
                         # Backstop: the engine answered, so it's no longer GIL-blocked
                         # loading. Clear the log-parsed loading state (the real task,