ExecuteOffloadResult(
self,
offloaded: bool,
response: ExecuteResponse,
preview_has_truncation_marker: bool = False
| Name | Type |
|---|---|
| offloaded | bool |
| response | ExecuteResponse |
| preview_has_truncation_marker | bool |
Result of BaseSandbox.execute_with_offload.
offloaded describes the capture mechanism and is kept off ExecuteResponse
(which an ordinary execute never sets).
Whether the output was left at the capture path.
When True, response.output holds only a preview of the output and the
full output lives at the capture path on the sandbox filesystem.
When False, response.output is the complete output.
The command result. response.truncated indicates the output hit the size cap.
Whether the preview contains a ... [N lines truncated] ... marker.
Set by the code that builds the preview. This is the only reliable way to know the marker was inserted, as the command output itself can contain a line that looks the same (scanning the text is unsafe).
This flag covers only the ... [N lines truncated] ... marker. A preview
may disclose other losses through its own in-band clip notices, which need
no flag because they carry their own explanation; see
_build_capture_execute_cmd in backends.sandbox for the wording.
When offloaded is False, this flag is always False.