using-git-worktrees

Creating an isolated git worktree involves branch-based directory selection that verifies `.worktrees/` is gitignored before setup. This prevents accidental commits of worktree artifacts into your main workspace.

Gitignore verification for project-local paths is required because without it, git worktree artifacts inside `.worktrees/` can be accidentally committed. If `git check-ignore` fails, you must update `.gitignore` and commit before proceeding.

The best way to run baseline tests in a new worktree is to use an automated setup that creates the branch, installs dependencies, runs tests, and stops to report failures. This ensures a clean test baseline before implementation begins.

Directory selection for git worktrees is deterministic: it prioritizes existing worktree directories, then reads `CLAUDE.md` preferences, and only asks you if no preference is found. This supports efficient parallel branch work.

Yes, automated worktree setup can auto-detect and run project-appropriate dependency installation. After creating the isolated branch workspace, it installs dependencies and executes baseline tests with pass/fail reporting.

When baseline tests fail during git worktree creation, the automated setup uses a stop-and-ask mechanism. It stops the process immediately to report the failures before allowing you to proceed with any feature implementation.