fix(voice): block new user turns immediately on update_agent() to prevent transition delay

livekit-agents/livekit/agents/voice/agent_activity.py

@@ -118,6 +118,7 @@ def __init__(self, agent: Agent, sess: AgentSession) -> None:
         self._started = False
         self._closed = False
         self._scheduling_paused = True
+        self._new_turns_blocked = False
 
         self._current_speech: SpeechHandle | None = None
         self._speech_q: list[tuple[int, float, SpeechHandle]] = []
@@ -758,6 +759,7 @@ async def _resume_scheduling_task(self) -> None:
             return
 
         self._scheduling_paused = False
+        self._new_turns_blocked = False
         self._scheduling_atask = asyncio.create_task(
             self._scheduling_task(), name="_scheduling_task"
         )
@@ -1401,7 +1403,7 @@ def _on_generation_created(self, ev: llm.GenerationCreatedEvent) -> None:
             # user_initiated generations are directly handled inside _realtime_reply_task
             return
 
-        if self._scheduling_paused:
+        if self._scheduling_paused or self._new_turns_blocked:
             # TODO(theomonnom): should we "forward" this new turn to the next agent?
             logger.warning("skipping new realtime generation, the speech scheduling is not running")
             return
@@ -1663,6 +1665,7 @@ def on_preemptive_generation(self, info: _PreemptiveGenerationInfo) -> None:
         if (
             not self._session.options.preemptive_generation
             or self._scheduling_paused
+            or self._new_turns_blocked
             or (self._current_speech is not None and not self._current_speech.interrupted)
             or not isinstance(self.llm, llm.LLM)
         ):
@@ -1699,7 +1702,7 @@ def on_end_of_turn(self, info: _EndOfTurnInfo) -> bool:
         # IMPORTANT: This method is sync to avoid it being cancelled by the AudioRecognition
         # We explicitly create a new task here
 
-        if self._scheduling_paused:
+        if self._scheduling_paused or self._new_turns_blocked:
             self._cancel_preemptive_generation()
             logger.warning(
                 "skipping user input, speech scheduling is paused",
@@ -1800,7 +1803,7 @@ async def _user_turn_completed_task(
             if self._rt_session is not None:
                 self._rt_session.interrupt()
 
-        if self._scheduling_paused:
+        if self._scheduling_paused or self._new_turns_blocked:
             logger.warning(
                 "skipping on_user_turn_completed, speech scheduling is paused",
                 extra={"user_input": info.new_transcript},
@@ -1833,7 +1836,7 @@ async def _user_turn_completed_task(
         elif self.llm is None:
             return  # skip response if no llm is set
 
-        if self._scheduling_paused:
+        if self._scheduling_paused or self._new_turns_blocked:
             logger.warning(
                 "skipping reply to user input, speech scheduling is paused",
                 extra={"user_input": info.new_transcript},

livekit-agents/livekit/agents/voice/agent_session.py

@@ -1245,6 +1245,11 @@ def update_agent(self, agent: Agent) -> None:
         self._agent = agent
 
         if self._started:
+            # immediately block the old activity from accepting new user turns
+            # during the transition window (before drain() formally pauses scheduling)
+            if self._activity is not None:
+                self._activity._new_turns_blocked = True
+
             self._update_activity_atask = task = asyncio.create_task(
                 self._update_activity_task(self._update_activity_atask, self._agent),
                 name="_update_activity_task",