Having pre commit hooks can be awesome and can suck at the same time. My work project implements pre commit hooks using husky and lint-staged. These hooks are for coding standards.

{
  "author": "Theorem.co",
  "scripts": {
    "fix:mylife": "$(npm bin)/eslint 'src/js/**' --cache --fix",
  },
  "devDependencies": {
    "husky": "4.2.5",
    "lint-staged": "10.1.6"
  },
  "husky": {
    "hooks": {
      "pre-commit": "lint-staged"
    }
  },
  "lint-staged": {
    "*.js": [
      "npm run fix:mylife"
    ],
    "*.rb": [
      "rubocop -a"
    ]
  }
}

This is great and it has really upped the quality of our code. The team likes it too but… yeah there’s always a but. Personally,I like to do a block of work then choreograph my commits, staging specific lines from files and then commit them. I do this back to back for many commits. It’s kind of like looking back and telling a story of how this came to be.

I’m sure you can imagine how difficult this could be if we are running eslint with a fix on each commit.

Enter the --no-verify or -n commit flag.

git commit -nm "Message for a no verify commit"

This will skip my commit hooks. Yeah its basically cheating but like a responsible engineer, I run all my format commands before I do this.