How should you name your branches on Git?
NicolasBrondinBernard
People who write incomprehensible commit messages, we see you!

Article published on 06/11/2020, last updated on 10/08/2026
When you use Git to version your code and you don't enforce a standardized naming convention, you very quickly lose the very point of using a version manager.
Because after all, how can you easily find your branches when half the names look like "fix", "another-fix" or "wtf"?
Branches
I distinguish three types of branches in a project: global branches, personal branches, and task branches.
Global branches
A global branch is a branch that is managed by only a very small number of developers and that has consequences on deployments. You should never code directly in these branches, all code must come from merge requests that have undergone a code review beforehand.
These branches are generally called: master, preprod, dev, ...
Personal branches
These branches are the main branches on which developers work autonomously on their respective sprint for the current dev version.
These branches are therefore all named: dev/[firstname], dev/[initials] or even dev/[firstname lastname].
Task branches
Sometimes one or more devs have to work on a task that falls outside their current workflow, say a bug fix on the current version, while their sprint is focused on the next version of the application.
So we create a branch specific to this task, based on the desired global branch, taking care to prefix the branch name with the type of task to be performed, for example:
- fix/user-signup
- doc/missing-route
- ux/onboarding-improvement
Careful, to keep a clean repository, the ideal is to delete each task branch once it has been merged with a global branch.
If you're interested in naming conventions for commits, go read our dedicated article: https://code-garage.fr/blog/comment-bien-ecrire-vos-messages-de-commit
Complete courses, exercises and certificates to really learn programming!
4.8 average rating
No comments yet