What is Husky and why use it?

Husky manages Git hooks so quality checks run automatically at commit and push time. It lets your whole team share the same pre-commit linting and commit-message rules without anyone remembering to run them manually.

What does lint-staged add over running ESLint directly?

lint-staged runs your linters only on the files staged for commit instead of the entire project. That makes the pre-commit hook fast and means it can auto-fix and re-stage just the files you touched.

How is Husky v9 different from older versions?

Husky v9 simplified hooks to plain shell scripts in the .husky folder with no boilerplate shebang or husky.sh source line required. You initialize with npx husky init and each hook file contains just the command to run.

What does the commit-msg hook check?

The commit-msg hook runs commitlint against your message, enforcing a convention like Conventional Commits. If the message does not match the rules, the commit is rejected so your history stays consistent and machine-parseable.

Can a teammate skip the hooks?

Hooks can be bypassed with git commit --no-verify, so they are a convenience gate, not a security control. For guaranteed enforcement, run the same lint, commitlint, and test checks again in your CI pipeline.

Husky Git Hooks Config Builder

Email me this result

Get this tool's output sent to your inbox, plus one useful tool a week. No spam, unsubscribe any time.

Add quality gates to every commit

Husky runs Git hooks for you, so linting, commit-message validation, and tests fire automatically at the right moment instead of relying on memory. This builder generates the three most useful hooks — pre-commit, commit-msg, and pre-push — along with the matching lint-staged config and the exact setup commands.

How it works

Husky installs a Git core.hooksPath pointing at a .husky folder. In Husky v9 each hook is a plain shell script containing just the command to run.

pre-commit runs lint-staged, which lints and formats only the files you staged. The lint-staged config maps a glob like *.{js,jsx,ts,tsx} to commands such as eslint --fix and prettier --write; matched files are auto-fixed and re-staged.
commit-msg runs commitlint --edit "$1", where $1 is the path to the file holding the in-progress commit message. If the message violates your convention, the commit aborts.
pre-push runs your test command so broken code never reaches the remote.

The setup commands install husky, lint-staged, and commitlint, run npx husky init, and write each hook file.

Tips and notes

lint-staged keeps pre-commit fast by only touching staged files.
Hooks can be bypassed with --no-verify, so re-run the same checks in CI.
Use npx --no-install in the commit-msg hook so it fails loudly if commitlint is missing rather than fetching it silently.
Keep the pre-push command lightweight; a full slow test suite belongs in CI.