Fix ci-bonsai-daily: guard on_depsgraph_update_caps during file load

on_depsgraph_update and on_depsgraph_update_caps are registered together
as persistent depsgraph handlers (bim/module/clip_box/__init__.py:50-52).
on_depsgraph_update guards with `if cls._file_loading: return`, but the
sibling on_depsgraph_update_caps did not, so a depsgraph tick during the
file-load window still ran it. Beyond the failing test, this can re-arm a
cap-rebuild bpy.app.timers callback in the exact load window _on_load_pre
cancels timers for, against regions whose GPU state is not yet wired.

Add the same _file_loading guard as the first check.

Verified in headless Blender:
test_clip_box.py::TestRefreshTimerLifecycle::test_depsgraph_update_no_op_while_loading
1 failed -> passed.

This change was made with the assistance of an AI tool.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

src/bonsai/bonsai/tool/clip_box.py

@@ -1270,6 +1270,17 @@ class ClipBox:
     def on_depsgraph_update_caps(cls, scene, depsgraph) -> None:
         """Depsgraph entry-point — guard, then delegate to the
         modal-aware debounce in :meth:`_handle_cap_tick`."""
+        # Same file-load danger window as on_depsgraph_update: a real
+        # depsgraph tick during load (between load_pre and the new file's
+        # first paint) must not re-arm a cap-rebuild timer. _on_load_pre
+        # already cancels any in-flight timer via _cancel_pending_cap_rebuild;
+        # without this gate, a depsgraph_update_post event firing later in
+        # the same load (Blender fires these while building the new file's
+        # scene) would immediately reschedule one via _handle_cap_tick,
+        # undoing that cancellation and re-arming against regions whose GPU
+        # state is not yet wired.
+        if cls._file_loading:
+            return
         if getattr(bpy.context, "screen", None) is None:
             return
         if cls._active_scene_props(scene) is None: