_prep_run_args() — langchain Function Reference

Architecture documentation for the _prep_run_args() function in base.py from the langchain codebase.

Function python AgentOrchestration ActionLogic calls 1 called by 2

Entity Profile

AgentOrchestration→ ActionLogic→ _prep_run_args() — langchain Function Reference

Dependency Diagram

graph TD
  9e451f25_97c2_f2c6_30eb_6fb62f743dc0["_prep_run_args()"]
  80cebff4_efbd_4311_85e9_de0dc81a7eee["base.py"]
  9e451f25_97c2_f2c6_30eb_6fb62f743dc0 -->|defined in| 80cebff4_efbd_4311_85e9_de0dc81a7eee
  ede8808b_c1b5_3ed7_439d_90ebc56dd263["invoke()"]
  ede8808b_c1b5_3ed7_439d_90ebc56dd263 -->|calls| 9e451f25_97c2_f2c6_30eb_6fb62f743dc0
  1d771513_400f_5264_361c_04bcb00f97ac["ainvoke()"]
  1d771513_400f_5264_361c_04bcb00f97ac -->|calls| 9e451f25_97c2_f2c6_30eb_6fb62f743dc0
  d8993587_9d39_6094_aba2_5ae5b188d613["_is_tool_call()"]
  9e451f25_97c2_f2c6_30eb_6fb62f743dc0 -->|calls| d8993587_9d39_6094_aba2_5ae5b188d613
  style 9e451f25_97c2_f2c6_30eb_6fb62f743dc0 fill:#6366f1,stroke:#818cf8,color:#fff

Relationship Graph

Source Code

libs/core/langchain_core/tools/base.py lines 1213–1247

def _prep_run_args(
    value: str | dict | ToolCall,
    config: RunnableConfig | None,
    **kwargs: Any,
) -> tuple[str | dict, dict]:
    """Prepare arguments for tool execution.

    Args:
        value: The input value (`str`, `dict`, or `ToolCall`).
        config: The runnable configuration.
        **kwargs: Additional keyword arguments.

    Returns:
        A tuple of `(tool_input, run_kwargs)`.
    """
    config = ensure_config(config)
    if _is_tool_call(value):
        tool_call_id: str | None = cast("ToolCall", value)["id"]
        tool_input: str | dict = cast("ToolCall", value)["args"].copy()
    else:
        tool_call_id = None
        tool_input = cast("str | dict", value)
    return (
        tool_input,
        dict(
            callbacks=config.get("callbacks"),
            tags=config.get("tags"),
            metadata=config.get("metadata"),
            run_name=config.get("run_name"),
            run_id=config.pop("run_id", None),
            config=config,
            tool_call_id=tool_call_id,
            **kwargs,
        ),
    )