Deleting a local branch

git branch -d BRANCH_NAME

Deleting a remote branch

git push origin --delete BRANCH_NAME


Stash all uncommitted files with a description

git stash save "DESCRIPTION"

Restore the most recent stash and remove it from the list of stashes

git stash pop

Restore the most recent stash and keep it in the list of stashes

git stash apply

Restore a specific stash

  1. List all available saved stashes

    git stash list
  2. Restore the stash

    • Restore the stash and remove it from the list of stashes

      git stash pop stash@{INDEX}
    • Restore the stash and keep it in the list of stashes

      git stash apply stash@{INDEX}


Undo the last local commit (will remove the last local commit, leaving your changes in place)

git reset HEAD~1

Undo the last public commit (like SVN revert; will make a new commit)

git revert HEAD

Rewriting history

Deleting files/directories from Git history

Use git-filter-repo:

  1. Do a fresh clone of the repo

  2. Use git-filter-repo, e.g.

    git-filter-repo --path PATH_TO_DELETE --invert-paths
  3. Add back the remote (git-filter-repo removes it), e.g.

    git remote add origin
  4. Force push to the remote

    git push --set-upstream origin main --force


Interactive rebasing/squashing

Warning: Rebasing should only be done on feature branches that nobody else is working on

  1. (Optional) Create a backup branch

    git checkout -b backup

    Then switch back to the branch you want to rebase

  2. Find the commit ID of the last commit you want to keep

  3. Start an interactive rebase

    git rebase -i COMMIT_ID


    git rebase -i 91b49d38
  4. Go through the list of commits and change pick as desired

    • squash is typically going to be the most useful option. It will take changes from that commit and add them to the previous commit.
      • It will make sure all changes get included without any conflicts and still gives you the opportunity to edit the commit message starting with the first pick as long as there’s at least one squash after it (not counting any fixup commits before or after the squash).
      • If you don’t want to change the previous commit, use squash or fixup with the next commit instead. Don’t use drop because it will result in conflicts since it will try to apply changes to pieces of the file that may have been dropped.
      • Since squash merges changes into the previous commit, it can’t be used for the first commit ID in a rebase.
    • fixup is really only useful for commits where you were testing or fixing typos and you’re absolutely certain you don’t want to use the commit message.
      • squash is typically still better because a pick followed by only fixup commits won’t give you the opportunity to edit the commit message.
    • Commit IDs will be changed starting with the first commit where you use something other than pick.
    • If you simply close the interactive rebase screen, nothing will be changed.
  5. Fix any conflicts as needed

    1. List the conflicted files

      git status
    2. Edit and save the conflicted files

    3. Use git add to mark the conflicted file as resolved. Ex:

      git add
    4. Resume the rebase

      git rebase --continue
  6. Push the rebased branch

    Use --force if this branch has already been pushed remotely

  7. Delete the backup branch

    git branch -d backup