fix(agent): bind finalize_turn at import time

The function-scoped import at the end of run_conversation loads
agent.turn_finalizer fresh from disk on the first turn that reaches it.
On a source/editable install whose checkout changed mid-session, that
pairs an old caller with a new callee at the exact seam where every turn's
work is persisted — the turn crashes on a signature mismatch after the
work is done. The lazy import was never cycle-forced: turn_finalizer
defers its own conversation_loop import.

agent/conversation_loop.py

@@ -86,6 +86,11 @@ from agent.retry_utils import (
     zai_coding_overload_retry_ceiling,
 )
 from agent.trajectory import has_incomplete_scratchpad
+# Bound at import time deliberately: a lazy end-of-turn import reads the
+# module fresh from disk, so a mid-session git switch/pull in a source
+# checkout loads a version-skewed finalizer and kills the finished turn.
+# Safe to hoist: turn_finalizer defers its own conversation_loop import.
+from agent.turn_finalizer import finalize_turn
 from agent.usage_pricing import estimate_usage_cost, normalize_usage
 from hermes_constants import PARTIAL_STREAM_STUB_ID
 from hermes_logging import set_session_context
@@ -7513,7 +7518,6 @@ def run_conversation(
     # Post-loop turn finalization extracted to agent/turn_finalizer.finalize_turn
     # (god-file decomposition Phase 1 step 4). Behavior-neutral: the assembled
     # result dict is returned exactly as before.
-    from agent.turn_finalizer import finalize_turn
     return finalize_turn(
         agent,
         final_response=final_response,