Job id to fetch outputs for (default: the most recent job).
NAME(optional)
Output name or output id to download. A name matching more than one output for this job is refused — pass the output id instead. Omitted: list the outputs.
Print the bare output list as a JSON array instead of a table (automation-friendly).
--limit N
Cap the bare list output (no NAME/–all) at N outputs (default 20 for the human-readable table; –json with no –limit returns the complete list instead, and –json WITH one says on stderr that it was bounded — machine output never truncates silently). 1..250. Does not change –all, which always downloads every output.
-o, --output OUTPUT
Output path. For one output, an existing directory or a path ending in a separator stores the server-named file there; any other path is the exact output file. For –all, this is a directory.
