{"version":"2.1.302","anchor":"worktree-resume-skips-the-resume-as-is-fast-path-when-a-re","canonical_anchor":"worktree-resume-skips-the-resume-as-is-fast-path-when-a-re","heading":"Worktree resume can skip reusing an existing worktree when no saved state is found","tier":"internal","area":"Worktrees","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.302\/e\/worktree-resume-skips-the-resume-as-is-fast-path-when-a-re","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.302","markdown":"### Worktree resume can skip reusing an existing worktree when no saved state is found\n\nWhen resuming into a worktree with no recorded state, Claude Code may now prepare it again rather than reuse it as is\n\n**Unclear.** The extra condition that must hold alongside a missing record is not known, so it is unclear when a worktree will no longer be reused.\n\n**What**\n\nA worktree is an extra working copy of a git repository. When Claude Code sets one up, it looks for a recorded state for it, meaning a saved commit. That lookup now reports either a recorded commit or that nothing was recorded. When nothing was recorded and a further condition holds, Claude Code no longer takes the path that reuses the existing worktree. Only the saved commit, rather than the whole record, is now passed on to the setup step.\n\n**Why**\n\nThis changes when an existing worktree is reused and when it is prepared again.\n\n- Area: Worktrees\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}