class documentation
class TemporalOpenAIRunner(AgentRunner): (source)
Constructor: TemporalOpenAIRunner(model_params)
Temporal Runner for OpenAI agents.
Forwards model calls to a Temporal activity.
| Method | __init__ |
Initialize the Temporal OpenAI Runner. |
| Async Method | run |
Run the agent in a Temporal workflow. |
| Method | run |
Run the agent with streaming responses. |
| Method | run |
Run the agent synchronously (not supported in Temporal workflows). |
| Instance Variable | model |
Undocumented |
| Method | _prepare |
Workflow-only validation and kwargs rewrite shared by run() and run_streamed(). |
| Instance Variable | _runner |
Undocumented |
async def run(self, starting_agent:
Agent[ TContext], input: str | list[ TResponseInputItem] | RunState[ TContext], **kwargs: Unpack[ RunOptions[ TContext]]) -> RunResult:
(source)
¶
Run the agent in a Temporal workflow.
def run_streamed(self, starting_agent:
Agent[ TContext], input: str | list[ TResponseInputItem] | RunState[ TContext], **kwargs: Unpack[ RunOptions[ TContext]]) -> RunResultStreaming:
(source)
¶
Run the agent with streaming responses.
Warning
Streaming inside Temporal workflows is experimental and may change in future versions.
Inside a workflow, model calls execute as the streaming model activity. The workflow consumes events via RunResultStreaming.stream_events() after each activity completes; external clients can subscribe to the configured stream topic to receive events as they arrive.
def run_sync(self, starting_agent:
Agent[ TContext], input: str | list[ TResponseInputItem] | RunState[ TContext], **kwargs: Any) -> RunResult:
(source)
¶
Run the agent synchronously (not supported in Temporal workflows).