DEV Community

Cole Walker
Cole Walker

Posted on

Conventional Commits, the Future of Git

Conventional commits are something which have revolutionized the way that I work, and made me re-think how I go about my git workflow.


Put simply, conventional commits is a standard for writing commit messages. Just like we have conventions for naming variables, or numbering release versions, there is a convention for writing commit messages.

Put even simpler, conventional commits tell you how to write your commit messages.

They're easier to write

There are many benefits to adopting conventional commits. But the first thing that an adopter notices is that they simplify the process of writing a commit. All developers are aware how difficult it can be to name things, and writing descriptive yet succinct commits can be even more difficult. By providing a format to follow, conventional commits take away a lot of the thinking work that is required when writing a commit message, and point us in a good direction.

They're easier to read

Secondly, they make it easier to read the history of a repository, for both humans and computers. The beauty of the convention is that it was designed for maintainers to easily be able to gloss over commits and get the gist of what each has done to the repository. When gitmojis are added into the fray, it makes it even easier.

Okay, so what are gitmojis?

Gitmojis are a convention (yes, another one) for attributing meaning to emojis in the context of development.

For example:

  • 🎨 = "Improve structure / format of the code."
  • ⚡️ = "Improve performance."
  • ✨ = "Introduce new features."

These emojis make it even easier to see what changes a commit made, because they're bright and instantly pop out to the reader. Even someone unfamiliar with the format can understand the general idea because of how humans associate meanings with images.

Tools can parse it

Conventions make it easy for automated tools to process commits. Tools exist to lint commits, help to write them, or even automate changelogs and versioning.

I'm sold, what's the convention?

<type>[optional scope]: <description>

[optional body]

[optional footer(s)]

This is a basic representation of the convention, written in Extended Backus–Naur Form. Further rules and explanations can be found on the conventional commits website.


The type of commit describes the type of work that occurs. You can think of this as being similar to the labels that are added to issues or stories.

The convention defines feat for features, and fix for bugfixes, but others can be added if necessary. For example, commitlint adds several more:

  • ci
  • chore
  • docs
  • feat
  • fix
  • perf
  • refactor
  • revert
  • style
  • test

Scope (optional)

Scope represents the area that you're performing the work on, and thus entirely depends on the project that is being worked on. Scope is usually surrounded by ( ). If a breaking change occurs, sometimes an ! is placed at the end of the scope to denote so.

For example, a web-application might have:

  • (database)
  • (dependencies)
  • (styling)


This is where the description of the commit goes. It is a brief statement describing the work that was done. Any gitmojis related to the changes should go at the beginning of this.

Body (optional)

This is where a longer description would go if necessary. If breaking changes were introduced with this commit, they must be explained either here or footer, in the format:
BREAKING CHANGE change description

Footer (optional)

Footers are entirely dependant on the project once again, but they must follow a specific format, "each footer MUST consist of a word token, followed by either a :<space> or <space># separator, followed by a string value." (From the conventional commits website)


(courtesy of the conventional commits website)

feat: ✨ allow provided config object to extend other configs

BREAKING CHANGE: `extends` key in config file is now used for extending other config files
refactor!: 💥 drop support for Node 6

BREAKING CHANGE: refactor to use JavaScript features not available in Node 6.
fix: ✏️ correct minor typos in code

see the issue for details

on typos fixed.

Reviewed-by: Z
Refs #133


Conventional commits are good and you should try them.

Top comments (7)

bchhun profile image
Bernard Chhun

Hey Cole,

I'm totally with you on this and I can't help but recommend using Semantic Versioning [1] + Semantic Release [2] in combination with Conventional Commits to automagically generate release notes/changelogs.

I just did a presentation yesterday for my colleagues at work on that subject [3]

The semantic release configuration is within the .releaserc file.

Running npm run release triggers a local release based on the commits history and the latest generated tag.


colewalker profile image
Cole Walker

This is awesome, thank you for sharing! I have an upcoming presentation at work regarding conventional commits + tooling, and I will absolutely bring Semver + Semantic Release up.

bloodgain profile image

My favorite part of this is requiring a "type" for the commit, which makes it even more clear that commits should do just one thing.

I also really like the idea of linting commit comments before accepting them, to ensure people are following good commit formatting rules as well as not leaving required parts out. It's crazy how many people ignore the community best practices, even though Tim Pope's rules are the Vim defaults in git.

For those not familiar with the accepted practice:

colewalker profile image
Cole Walker

Thanks for sharing, that was an awesome article!

colewalker profile image
Cole Walker

That's exactly the conventional commits extension that I would recommend. It's an excellent choice.

Thank you for sharing the article, I greatly appreciate it!

hasii2011 profile image
Humberto A Sanchez II

Can’t wait to start using this technique. Typically, I just put the issue URL and a small comment. Like the article says using a common format makes it machine parseable.

axelsomerseth profile image
Axel Somerseth Cordova

Thanks! Nice post!