Every developer has done it: you start work on a feature, make a bunch of changes, commit some of them — and then realise the branch was created off master when it should have been off develop, or vice versa.
The naive fix is to copy your changes manually, create the correct branch, and paste everything back. There is a better way.
The command
Create the correct branch from the right base, check it out, then pull the changes from the faulty branch:
git checkout feature/faulty-branch -- .
Run this while you are on feature/correct-branch. Git will copy all tracked file changes from feature/faulty-branch into your working tree, staged and ready to commit. No manual copying, no risk of missing a file.
What it actually does
git checkout <branch> -- <path> checks out the contents of <path> from <branch> into your current working tree. Using . means “everything from the repo root”. The changes land as staged modifications — you can review them with git diff --cached before committing.
Note that this copies file state, not commit history. You get the changes but not the individual commits. If you need to preserve commit history, git rebase --onto is the right tool, but for most day-to-day “wrong branch” mistakes this is faster and less error-prone.
The full workflow
# 1. Create the correct branch from the right base
git checkout develop
git checkout -b feature/correct-branch
# 2. Pull changes from the faulty branch
git checkout feature/faulty-branch -- .
# 3. Review and commit
git diff --cached
git commit -m "your message"
# 4. Clean up the faulty branch when ready
git branch -d feature/faulty-branch
This has saved me a surprising amount of time over the years. Simple, fast, and easy to remember.
