Bug: Agent callback - on_llm_start documentation

[michelebelmonte]

Expected Behaviour

The inline documentation of the on_llm_start method should contain only the actual parameters of the method.

Current Behaviour

A "messages" parameters is mentioned, even though it is not part of the method signature

Code snippet

async def on_llm_start(
        self,
        name: str,
        payload_input: Any,
        run_id: Optional[UUID] = None,
        tags: Optional[list[str]] = None,
        metadata: Optional[dict[str, Any]] = None,
        **kwargs: Any,
    ) -> Any:
        """
        Callback method that runs when an llm starts processing.

        This method is called at the beginning of an llm's execution, providing information
        about the llm session and its context.

        Parameters:
            self: The instance of the callback handler class.
            agent_name: Name of the agent that is starting.
            payload_input: Dictionary containing the agent's input.
            messages: List of message dictionaries representing the conversation history.
            run_id: Unique identifier for this specific agent run.
            tags: Optional list of string tags associated with this agent run.
            metadata: Optional dictionary containing additional metadata about the run.
            **kwargs: Additional keyword arguments that might be passed to the callback.

        Returns:
            Any: The return value is implementation-dependent.
        """
        pass

Possible Solution

Remove this line:
messages: List of message dictionaries representing the conversation history.

also this "agent_name" should be changed to "name"

Steps to Reproduce

Read the inline documentation for the on_llm_start method