"When Agent A writes code in one file that breaks Agent B's assumptions in another file." Each agent runs in a separate worktree or cloned copy of the repo, in

"When Agent A writes code in one file that breaks Agent B's assumptions in another file." Each agent runs in a separate worktree or cloned copy of the repo, independently. When the task is done, a PR is opened. The issue you mentioned gets caught during merge conflicts. No different than a "real" dev team?