Temporarily shelve uncommitted work-in-progress changes and restore them later without committing

Documented steps

1. Save your current modifications and revert the working directory to HEAD: `git stash push` (bare `git stash` is equivalent). Add a descriptive label with `git stash push -m "<message>"` since stashes are otherwise auto-named 'WIP on <branch>...'.
2. To also stash untracked files (which are ignored by default), use `git stash push -u` / `--include-untracked`; to also sweep in ignored files, use `-a`/`--all`.
3. To stash only specific files, pass a pathspec after `--`: `git stash push -- <path1> <path2>`; this records the modified state only for matching files and rolls back only those files to HEAD, leaving the rest of the working tree untouched.
4. List saved stashes with `git stash list` — entries are named `stash@{0}` (most recent), `stash@{1}`, etc., and can also be referenced by log-like syntax such as `stash@{2.hours.ago}`. Inspect one with `git stash show -p stash@{0}` to view it as a patch.
5. Restore the most recent stash and remove it from the list with `git stash pop`; to reapply a specific one, `git stash pop stash@{n}`. If you want to reapply without removing it from the list (e.g. to apply to multiple branches), use `git stash apply [stash@{n}]` instead.
6. If `pop`/`apply` produces conflicts, the stash entry is deliberately NOT dropped so you don't lose it; resolve the conflicts by hand, `git add` the files, then run `git stash drop stash@{n}` manually to remove it once you're satisfied.
7. Once a stash is no longer needed, delete it explicitly with `git stash drop [stash@{n}]`, or remove all stashes at once with `git stash clear`.