Lighthouse Tracker

Run Lighthouse against your website, parse the JSON result for only those metrics that are useful, and return the result. From there you can track whether your key metrics are improving (or not).

Motivation

You can read about the thinking behind this here: but basically, we wanted a way to run Lighthouse against our site whenever we released a build via CircleCI.

Installation

Either npm:

npm install lighthouse-tracker

Or yarn:

yarn add lighthouse-tracker

Commands

lighthouse.run(url)

This returns a Promise with data in the format below. It's up to you how you want to store/display that data e.g. save to a database, POST to a service like Datadog.

[
  {
    "date": "2019-03-01T07:57:59.368Z",
    "url": "https://anucreative.com/",
    "id": "first-contentful-paint",
    "title": "First Contentful Paint",
    "rawValue": 1118.575,
    "displayValue": "1.1 s",
    "score": 1
  },
  {
    "date": "2019-03-01T07:57:59.368Z",
    "url": "https://anucreative.com/",
    "id": "first-meaningful-paint",
    "title": "First Meaningful Paint",
    "rawValue": 1251.874,
    "displayValue": "1.3 s",
    "score": 1
  },
  {
    "date": "2019-03-01T07:57:59.368Z",
    "url": "https://anucreative.com/",
    "id": "speed-index",
    "title": "Speed Index",
    "rawValue": 1655,
    "displayValue": "1.7 s",
    "score": 1
  }
  ...
]

It currently returns the following metrics:

[
  'first-contentful-paint',
  'first-meaningful-paint',
  'speed-index',
  'network-requests',
  'interactive',
  'total-byte-weight',
  'uses-webp-images',
  'uses-optimized-images',
  'uses-responsive-images'
]

Example (simple)

1. Create a runner file

// lighthouse-runner.js
 
const lighthouse = require('lighthouse-tracker')
 
const saveData = data => {
  // Post to your favourite data store
}
 
// Run
lighthouse
  .run("https://welcometothejungle.com")
  .then(saveData)
  .catch(console.error)

2. Run your file from npm scripts

// package.json
 
{
  "scripts": {
    "lighthouse": "node ./lighthouse-runner.js",
  }
}

Example (more involved)

1. Create a runner file that will read the URL from the command line arguments

// lighthouse-runner.js
 
const path = require('path')
const dogapi = require('dogapi')
const lighthouse = require('lighthouse-tracker')
const parseArgs = require('minimist')
 
require('dotenv')
 
// Get URL from arguments
const argv = parseArgs(process.argv.slice(2))
const url = argv._[0]
if (!url) {
  throw Error('No URL provided')
}
 
// Post to Datadog
const postToDataDog = metrics => {
  dogapi.initialize({
    api_key: process.env.DATA_DOG_API_KEY,
    app_key: process.env.DATA_DOG_APP_KEY
  })
  return dogapi.metric.send_all(metrics, (err, res) => {
    if (err) {
      throw new Error(err)
    }
    console.dir(res)
  })
}
 
// Run
lighthouse
  .run(url)
  .then(postToDataDog)
  .catch(console.error)
 

2. Set up the npm task (same as first example)

// package.json
 
{
  "scripts": {
    "lighthouse": "node ./lighthouse-runner.js",
  }
}

3. Call the lighthouse npm task from Circle with a URL

// .circleci/config.yml
 
jobs: 
  lighthouse:
    steps:
      - run: yarn lighthouse https://$AUTH_USER:$[e-mail]
      - store_artifacts:
          path: reports/lighthouse 

Todo

• Tests
• Accept more Lighthouse arguments (e.g. throttle, list of metrics)
• Multiple passes taking median results