doctor

gitto doctor [<name>] [--json]

Reports what still points outside a clone after the copy: symbolic links into the canonical, virtual environments holding the old path, submodule history that did not come along, worktrees the clone hosts for someone else, and settings files in fixed places that name the clone by path.

With no name it reports the canonical as well. A canonical hosts worktrees like any other checkout, and it is the one nobody thinks to check before renaming it.

$ gitto doctor
storefront  canonical
  it hosts no worktree that lives outside it
auth-fix
  nothing git can see points at the canonical

What it can see

Everything git records, and the two cases that bite most often: a symbolic link whose target resolves outside the clone, and a virtual environment whose activate exports the path it was built at.

For every worktree a checkout hosts, at the top level and at each submodule depth, it reads that worktree's own gitdir marker and says when the two disagree: the directory is gone, the marker names a gitdir that no longer exists, or the marker resolves into a different checkout. That last case is the quiet one. A marker holds an absolute path, so renaming a checkout frees that address for whatever moves in next, and the marker goes on resolving to a repository that never registered it. Git reports nothing on either side.

It also reads a short list of settings files that live at a fixed address, since a path written there survives every move and reports nothing when it breaks: ~/.ssh/config, ~/.gitconfig, /etc/fstab, launch agents under ~/Library/LaunchAgents, unit files under ~/.config/systemd/user, and the calling user's crontab.

What it cannot see

Anything else outside git that names a path: an editor workspace, a build cache, a shell sitting in the directory, a script in a place nobody thought to look. The fixed list above is a sample, so a clean report means the references git knows about are in order and those few files say nothing. It does not mean the directory is safe to delete.