You are probably here because you found this address in your server
logs. This page explains who we are, why we visited, and how to stop us.
The short version
We fetch a small number of pages from a website when that site's owner (or
someone they have authorised) orders an accessibility audit of it. We read
pages only — we never log in, never submit forms, never buy anything, and
never touch anything behind a password.
How to identify our traffic
Our crawler always sends this User-Agent, with a unique scan ID per job:
If you contact us with that scan ID, we can tell you exactly which order the
traffic belonged to and when it ran.
Our scans run on Render
infrastructure in the US Oregon region and egress from Render's published
outbound IP ranges for that region. Those ranges are shared with other Render
customers, so an IP match alone does not identify us — the User-Agent and scan
ID do. We do not scan from residential connections, and we do not use
proxy rotation.
What we do
Request up to 25 pages of a site, one at a time.
Wait at least two seconds between pages, and honour
Crawl-delay in your robots.txt up to 15 seconds.
Read and obey robots.txt before fetching anything else.
Prefer your sitemap.xml over following links.
Render each page in a headless browser and run the open-source
axe-core rule set
against it, then capture a cropped screenshot of failing elements.
Stop immediately and completely on the first
429, 401, 403 or 503.
We do not retry into a block.
Give up rather than proceed if we detect a bot-protection challenge page.
What we never do
Log in, or accept credentials, cookies or session tokens — even if offered.
Submit forms, place orders, or change any state on your site.
Disguise ourselves. We do not spoof the User-Agent, rotate proxies, or use
stealth plugins to evade bot protection. If you block us, we stay blocked.
Scan .gov, .mil, .edu or equivalent
public-sector domains at all.
Sell, share, or republish anything we read. Findings go only to the
customer who ordered the audit of that site.
How to stop us
Add this to your robots.txt and we will not come back:
User-agent: AIgentechA11yBot
Disallow: /
We check robots.txt at the start of every job and cache it for no
more than 24 hours. If you would rather not wait, email us (below) and we will
add your domain to a permanent do-not-scan list.
What we keep
The scan results (which rules failed, on which elements) and cropped
screenshots of the failing elements.
A request log: how many requests we made, how long the scan took, and the
HTTP statuses we saw.
Everything is deleted after 30 days, or sooner on request.
We strip page text, email addresses and phone numbers out of stored markup,
and we skip screenshots on pages that look like staff, team or contact pages.
Contact
Abuse reports, opt-out requests, and questions about a specific scan ID:
abuse@aigentech.com. A person reads this address and we will
respond. If our traffic caused you a problem, tell us — we would rather fix it
than have you find out from your logs again.