using-git-worktrees

What git worktrees solve

git worktree lets one repo have multiple checked-out branches in separate directories simultaneously. Without worktrees, the choices are:

1. Branch-switch with stash — git stash your work, git switch other-branch, do work, git switch back, git stash pop. Loses the running build artifacts, IDE state, open files. Stash conflicts are a recurring sharp edge.
2. Multiple clones — git clone the repo into a second directory. Costs disk space (full .git/ × N) and de-syncs (each clone has its own remote refs, branch list, hooks).
3. Just don't context-switch — works until reality forces a switch (production bug while feature is in flight).

git worktree is a fourth option: one .git/ shared across N checkouts in N directories, each on its own branch. Cheap to create (no full clone); cheap to destroy (just delete the directory + git worktree prune); fully synced (one remote, shared hooks, shared object store).

When to use