temporalio.contrib.langgraph._langgraph_config

module documentation
(source)

LangGraph configuration management for Temporal workflows.

Function get_langgraph_config Get the current LangGraph runnable config as a serializable dict.
Function set_langgraph_config Restore a LangGraph runnable config from a serialized dict.
Function strip_runnable_config Return a serializable subset of a RunnableConfig.
def get_langgraph_config() -> dict[str, Any]: (source)

Get the current LangGraph runnable config as a serializable dict.

def set_langgraph_config(config: dict[str, Any]) -> Runtime: (source)

Restore a LangGraph runnable config from a serialized dict.

Returns the reconstructed Runtime so callers can re-inject it into the user function's kwargs without needing to know the configurable layout.

def strip_runnable_config(config: RunnableConfig | None) -> RunnableConfig: (source)

Return a serializable subset of a RunnableConfig.

LangGraph injects the active RunnableConfig into user functions as a config kwarg. The full object holds non-serializable things (callbacks, checkpointer/store/cache handles, pregel send/read callables) that can't cross an activity boundary, so we keep only primitive fields and the serializable subset of configurable.