Skip to main content

The geta11y crawler

geta11y checks websites for accessibility barriers. To do that, our crawler reads the pages of a website the way a visitor's browser does.

When it visits a website

  • When someone has asked geta11y to check that website.
  • Once a quarter for our published research: the home page of online shops, one page per shop.

How to recognise it

While it finds the pages of a website, the crawler names itself in every request.

User-Agent
AccessibilityAgent/1.0 (WCAG Auditor)

The pages it then checks are loaded in a Chrome browser, because a check has to see the page as a visitor sees it. Those requests carry Chrome's own User-Agent.

Requests to the website being checked are signed (Web Bot Auth), from the crawler and from the browser alike. The public keys to verify the signature are published here.

Public keys
https://geta11y.com/.well-known/http-message-signatures-directory

What it does and does not do

  • It follows the website's robots.txt when it finds pages, including a Crawl-delay.
  • It loads pages slowly, and it waits when a website answers that it is receiving too many requests.
  • It never places an order.
  • In a paid audit of an online shop it adds one product to a cart to reach the cart and checkout pages, and it opens those two pages even where robots.txt closes them.

Keeping it out

Add this rule to your robots.txt and no page of the website is checked.

User-agent: AccessibilityAgent
Disallow: /

Contact

Questions about a visit from our crawler are answered at this address.

info@geta11y.com