feat: add extra_create_args to AssistantAgent for model client custom…

…ization

python/packages/autogen-agentchat/src/autogen_agentchat/agents/_assistant_agent.py

@@ -64,6 +64,7 @@ class AssistantAgentConfig(BaseModel):
     system_message: str | None = None
     reflect_on_tool_use: bool
     tool_call_summary_format: str
+    extra_create_args: Mapping[str, Any] | None = None
 
 
 class AssistantAgent(BaseChatAgent, Component[AssistantAgentConfig]):
@@ -147,6 +148,7 @@ class AssistantAgent(BaseChatAgent, Component[AssistantAgentConfig]):
             Available variables: `{tool_name}`, `{arguments}`, `{result}`.
             For example, `"{tool_name}: {result}"` will create a summary like `"tool_name: result"`.
         memory (Sequence[Memory] | None, optional): The memory store to use for the agent. Defaults to `None`.
+        extra_create_args (Mapping[str, Any] | None, optional): Additional arguments to pass to the model client during the create method call. Defaults to `None`.
 
     Raises:
         ValueError: If tool names are not unique.
@@ -271,6 +273,7 @@ def __init__(
         reflect_on_tool_use: bool = False,
         tool_call_summary_format: str = "{result}",
         memory: Sequence[Memory] | None = None,
+        extra_create_args: Mapping[str, Any] | None = None,
     ):
         super().__init__(name=name, description=description)
         self._model_client = model_client
@@ -337,6 +340,7 @@ def __init__(
         self._reflect_on_tool_use = reflect_on_tool_use
         self._tool_call_summary_format = tool_call_summary_format
         self._is_running = False
+        self._extra_create_args = extra_create_args or {}
 
     @property
     def produced_message_types(self) -> Sequence[type[ChatMessage]]:
@@ -384,7 +388,7 @@ async def on_messages_stream(
         # Generate an inference result based on the current model context.
         llm_messages = self._system_messages + await self._model_context.get_messages()
         model_result = await self._model_client.create(
-            llm_messages, tools=self._tools + self._handoff_tools, cancellation_token=cancellation_token
+            llm_messages, tools=self._tools + self._handoff_tools, extra_create_args=self._extra_create_args, cancellation_token=cancellation_token
         )
 
         # Add the response to the model context.
@@ -465,7 +469,9 @@ async def on_messages_stream(
         if self._reflect_on_tool_use:
             # Generate another inference result based on the tool call and result.
             llm_messages = self._system_messages + await self._model_context.get_messages()
-            model_result = await self._model_client.create(llm_messages, cancellation_token=cancellation_token)
+            model_result = await self._model_client.create(
+                llm_messages, extra_create_args=self._extra_create_args, cancellation_token=cancellation_token
+            )
             assert isinstance(model_result.content, str)
             # Add the response to the model context.
             await self._model_context.add_message(AssistantMessage(content=model_result.content, source=self.name))
@@ -540,6 +546,7 @@ def _to_config(self) -> AssistantAgentConfig:
             else None,
             reflect_on_tool_use=self._reflect_on_tool_use,
             tool_call_summary_format=self._tool_call_summary_format,
+            extra_create_args=self._extra_create_args,
         )
 
     @classmethod
@@ -555,4 +562,5 @@ def _from_config(cls, config: AssistantAgentConfig) -> Self:
             system_message=config.system_message,
             reflect_on_tool_use=config.reflect_on_tool_use,
             tool_call_summary_format=config.tool_call_summary_format,
+            extra_create_args=config.extra_create_args,
         )