More efficient Grunt workflows

Useful tips for development with Grunt

Most of my projects are fairly similar: I need to concatinate and compress a bunch of js files together; compile sass and fire-up a local server to test and preview.. Grunt.js is brilliant for this and I cannot tell you how much it has improved my workflow since I started using it this year. Along the way I have found a few time-savers that have greatly improved how I use Grunt so I thought I’d share them here with you.

Dynamically load all Grunt tasks

If you use a bunch of Grunt tasks for your project you’ll know that it is a pain, albeit slight, to individually register each task, especially when adding new packages. Why not register them all at once so that every time you add a new package you do not have to do anything except install it? Well load-grunt-tasks by Sindre Sorhus does just that.

// Use this:

// Instead of this:

See how it’s done in Kickoff’s Gruntfile

Check and find updates for your NPM dependencies

Have you ever found it hard to know if grunt-contrib-watch has been updated recently? What would happen if you wanted to check all your project’s dependencies and then install them? It would take ages. npm-check-updates takes care of this for you; it actually does the check and can even update your package.json with any changes so you all you now need to do is run npm-check-updates -u && npm install. I suggest installing the package globally (sudo npm install npm-check-updates -g) so that you can always use it from any directory.

To make this even easier to use I created a bash alias:

alias npmui="npm-check-updates -u && npm install"

Add this to your ~/.bash-profile and you’re good to go.

Use variables to avoid repetition

Typically there will be about two or three places within my Gruntfile.js that I declare the same group of files that will be used by certain tasks so to avoid repetition I create an array that lists each file which I use throughout my Gruntfile (I also use simple variables to specify file or directory names).

A good example of this might be javascripts used by jshint, uglifyjs and my watch task. Where each of these tasks specify a src key, I provide a variable as their value. See below for a simplified example of this or have another look at Kickoff’s Gruntfile for a more extensive example.

 * Specify which js files you want to include
var jsFileList = [

 * Specify your output directory
var distDir = 'js/dist/';

 * Specify the name of your compiled JS file
 * which will be placed in the directory you define above
var jsFile = 'app.min.js';

 * Project configuration
  pkg: require('./package'),

   * JSHint
   * Manage the options inside .jshintrc file
  jshint: {
    all: jsFileList,
    options: {
      jshintrc: '.jshintrc',

   * Uglify
   * Minifies and concatinates your JS
   * Also creates source maps
  uglify: {
    options: {
      sourceMap: distDir + jsFile + '.map',
      sourceMappingURL: jsFile + '.map',

     * Use the array at the top of this file to specify which js files you include
    js: {
      src: jsFileList,
      dest: distDir + jsFile,

Create local server and watch for changes

Grunt connect is extremely useful for creating a local server, but having it running means that you can’t run grunt watch at the same time. To overcome this problem, register a new Grunt task called serve that runs grunt watch and grunt connect. See below:

grunt.registerTask('serve', ['connect', 'watch']);

Use grunt serve to run the task; see how this was used in Kickoff’s Gruntfile and in particular lines 171-184.


These are just a few that I use regularly but I’m sure there are more. Please leave a comment with any suggestions or other useful tips you may have.

If you haven’t used Grunt before and are looking to improve your workflow I highly recommend checking it out, see for more info.