OptionalerrorThe envelope's error.message, present when status === "failed".
OptionalincompleteThe envelope's incomplete_details.reason, present when
status === "incomplete" (e.g. max_output_tokens,
content_filter). Lets a completion-side consumer classify an
incomplete response without re-deriving it.
OptionaloutputParsed text output, present when status === "completed" and
the response carried a message output item.
The OpenAI response id that was retrieved.
Current status of the stored response.
OptionaltokenToken usage reported by OpenAI, when available.
The structured result of a
retrieveResponsecall. All fields exceptstatusandrawResponseIdare absent for non-terminal or failed responses.