I’m considering using Git worktrees so two Claude Code sessions can tackle separate tasks at the same time. The goal is to keep each session’s file changes isolated, while still letting both work from the same project history.
Has anyone tried this in a real repository? I’m curious whether Claude Code stays reliably within the worktree it was started in, and whether there are any gotchas with shared configuration, branches, or test processes. I’d rather catch the awkward parts before I let parallel sessions loose on a busy codebase.
Can you answer this question?
Write Answer0 Answers