redis_openai_agents.MiddlewareStack

class MiddlewareStack(model, middlewares)

Compose middleware around an OpenAI Agents SDK Model.

Example:

base = OpenAIResponsesModel(model="gpt-4o")
stack = MiddlewareStack(
    model=base,
    middlewares=[SemanticCacheMiddleware(cache), MetricsMiddleware(metrics)],
)
result = await Runner.run(agent, "Hello", model=stack)

Parameters:

• model (Any) – The inner Model instance that performs the actual LLM call.

• middlewares (Sequence[AgentMiddleware]) – Sequence of middleware applied outermost-first.

__init__(model, middlewares)

Parameters:

• model (Any)

• middlewares (Sequence[AgentMiddleware])

Return type:

None

Methods

__init__(model, middlewares)
close()	Release any resources held by the model.
get_response(system_instructions, input, ...)	Run the middleware chain then the underlying model.
get_retry_advice(request)	Return provider-specific retry guidance for a failed model request.
stream_response(system_instructions, input, ...)	Delegate streaming directly to the inner model.

Attributes

inner	The wrapped Model instance.
middlewares	Applied middlewares, in outer-to-inner order.