this is also the ulitimate cli coding agent benchmark, by the way. give it a time limit (12 hrs?) and a cost limit ($1k?) and let claude code, codex, amp cli, etc etc duke it out. judges just use the final result and see who wins.
X (formerly Twitter) ·
this is also the ulitimate cli coding agent benchmark, by the way. give it a time limit (12 hrs?) and a cost limit ($1k?) and let claude code, codex, amp cli, etc etc duke it out. judges just use the final result and see who wins.