Some old documentation about the goals of htmllint follows:

htmllint verifies that html files match certain specifications. This functionality is useful for almost anyone who writes html; a few more specific cases are given below.

• A beginner to html wishes to verify that his or her test webpages are correct and follow good practice. He will use JavaScript to call htmllinter.
• The leader of a programming team at a company wants to maintain and enforce a set of standards for the html documents the team writes. They use Grunt for automation of tasks at will.
• A hobbyist web programmer or freelancer wants to make sure his or her html pages follow current standards, using a highly customizable tool to account for his or her personal style decisions. He uses the command line to access npm modules for his work.

• Follow a specified format for naming rules.
• It should be possible to configure any rule.
• Configurable rules should be highly flexible - meaning everything we can provide, we should provide.

1. The software will be compatible with Grunt and available via npm.
2. Verification of basic HTML5 compliance (e.g. matched angle brackets and tags).
3. Verification of essential good practice like providing a source for images and quoted attribute values.
4. Verification of style issues like mixed tabs and spaces.
5. Configurability: it should be possible to disable or modify any rule.
6. Ease of configurability: the above should be simple, intuitive, and consistent.
7. Completeness of output: Output should include full debugging information and give the rule violated as well as the column and line number.
8. Ease of reading output: Output should be concise but intuitive.
9. Performance: Should finish small pages (under 100 KB) in under 1 second.
10. Extendability: Users should be able to add custom tests using javascript and without modifying the htmllint source.