Enhance greeting context management in Brain classes

- Introduce greeting context handling in BaseBrain and WorkflowBrain to manage assistant greetings effectively.
- Implement prepare_greeting_context method to add greeting messages to the local context while preserving playback order.
- Update pipeline event handling to ensure greeting timestamps are maintained until the client is ready.
- Enhance tests to verify the correct behavior of greeting context management in various scenarios.

backend/services/pipecat/pipeline_events.py

@@ -32,6 +32,7 @@ def bind_cascade_pipeline_events(
 
     pending_text_inputs: list[str] = []
     greeting_transcript_sent = False
+    greeting_timestamp = ""
 
     async def queue_transcript(role: str, content: str, timestamp: str) -> None:
         if not content:
@@ -122,11 +123,16 @@ def bind_cascade_pipeline_events(
         nonlocal greeting_transcript_sent
         if greeting and not greeting_transcript_sent:
             greeting_transcript_sent = True
-            await queue_transcript("assistant", greeting, time_now_iso8601())
+            await queue_transcript(
+                "assistant",
+                greeting,
+                greeting_timestamp or time_now_iso8601(),
+            )
         await brain.on_client_ready()
 
     @transport.event_handler("on_client_connected")
     async def on_client_connected(_transport, _client):
+        nonlocal greeting_timestamp
         if vision_enabled:
             try:
                 vision_state["client_id"] = get_transport_client_id(
@@ -140,8 +146,11 @@ def bind_cascade_pipeline_events(
             except Exception as exc:  # noqa: BLE001 - media availability is optional
                 logger.warning(f"视觉理解摄像头捕获初始化失败: {exc}")
         if greeting:
+            # Preserve the actual playback order. The transcript is delivered
+            # later on client-ready, but the preview sorts by this timestamp.
+            greeting_timestamp = greeting_timestamp or time_now_iso8601()
             if brain.spec.owns_context:
-                context.add_message({"role": "assistant", "content": greeting})
+                brain.prepare_greeting_context(greeting, context)
             await worker.queue_frame(
                 TTSSpeakFrame(greeting, append_to_context=False)
             )