SeekyoBot

Identification

SeekyoBot/0.1 (+https://seekyo.aipinto.com/bot)

Every request carries that header. There is no second user agent, and no attempt to look like a browser.

Why it was on your site

A customer asked a question and pointed Seekyo at a starting page. Seekyo followed links from there looking for passages that answer that specific question. It was not indexing your site, not building a general purpose copy of it, and not crawling it on a schedule: the request happened because somebody asked, and only the pages reached by following links from their starting point were fetched.

Text from a page can end up as a quoted passage in that customer's result. Assets are never fetched: no images, scripts, stylesheets or fonts.

What it respects

• robots.txt, per RFC 9309. A group naming SeekyoBot applies to it, and so does the * group when no group names it.
• An unreadable robots.txt means stop. If the file returns a server error, or the fetch fails, that origin is treated as completely disallowed and nothing on it is fetched. This is the stricter reading in the standard, chosen deliberately.
• A missing robots.txt does not. A 404 means the same as it does anywhere else: there are no stated rules, so nothing is disallowed by them.
• noindex and nofollow, whether declared in a robots meta tag or in an X-Robots-Tag response header. A noindex page's text is never returned to the customer. A nofollow page's links are never followed.
• Crawl-delay, honoured up to a ceiling of ten seconds. A larger declared delay is capped rather than obeyed indefinitely.
• Rate limiting: one request in flight per origin at a time, spaced by at least 250 milliseconds with jitter added on top, never subtracted.
• Response size, capped, so a large or endless response cannot be read into memory in full.

What it never does

• No JavaScript execution, no headless browser, no rendering of any kind.
• No cookies, no sessions, and no authentication: it only requests what an anonymous visitor could request.
• No CAPTCHA solving, no fingerprint or user agent spoofing, no stealth behaviour.
• No form submission, no paywall circumvention, no credential discovery.
• No POST, PUT or DELETE. It only sends GET.
• No crawling of a page it was told not to use, and no following of links from such a page.

How to block it

Add this to the top of your robots.txt:

User-agent: SeekyoBot
Disallow: /

# or, to slow it down rather than stop it:
User-agent: SeekyoBot
Crawl-delay: 10

Blocking by user agent at your edge works too, as does blocking our egress address. If you would rather we stopped entirely, write to us and we will add your host to a deny list that applies across every customer.

Contact

aipinto.dev@gmail.com. Please include the hostname and a timestamp if you are reporting a problem rather than asking a question.