Last updated October 18, 2026

SeiBot

SeiBot is the scanner behind Sei's Site health. It reads a website's public pages the way search engines and AI crawlers see them, and reports what could stop a page from being indexed, quoted or understood: status codes and redirects, robots rules and canonical tags, structured data, sitemaps and how much text the HTML carries.

How to recognise it

Every request SeiBot makes carries this User-Agent:

SeiBot/1.0 (+https://getsei.vercel.app/bot)

Which pages it reads

Only sites someone added to a Sei workspace, and only a bounded set of their pages: robots.txt, the home page, the articles published from Sei, and up to 50 more pages from the site's sitemap (it fetches at most 20 sitemap files per scan). It never follows links to discover other pages, never submits forms, never signs in, sends no cookies and doesn't run JavaScript.

How often, and how politely

  • A scan starts when someone clicks Scan now, once a day for sites connected to Sei, or after an article is published (then usually only that page). At most 3 scans of a domain start in any hour, however many Sei workspaces added it, and only one runs at a time.
  • At most 2 requests at a time to a site, each waiting 1 second between its requests, with a 15-second timeout and the first 3 MB of each page read.

Blocking or allowing it

SeiBot obeys robots.txt (RFC 9309): the rules in a User-agent: SeiBot group, or else those for User-agent: *. It doesn't fetch a page those rules disallow, and if they disallow the home page it doesn't scan the site at all. For example:

User-agent: SeiBot · Disallow: /

If your firewall blocks SeiBot instead, the Sei workspace that added your site sees that the scan was blocked. Questions or a request to stop: contact us.

Allowing SeiBot through a firewall

A firewall or bot check that challenges SeiBot (Cloudflare's “Just a moment…” page, a 403 or a captcha) stops Sei from reading the site: onboarding can't tell its platform, and Site health scans come back blocked. The same check stops Sei's server from reaching the WordPress REST API, so connecting WordPress, importing and publishing fail too. Sei's server can't solve a challenge meant for a browser; the firewall has to let it through. Every request it sends carries SeiBot in its User-Agent (the WordPress connection sends Sei-WordPress/1.0 (SeiBot; +https://getsei.vercel.app/bot)), so one rule on that word covers both.

Cloudflare. SeiBot isn't one of Cloudflare's verified bots, so Cloudflare doesn't let it in on its own.

  • Add a WAF custom rule: in the Cloudflare dashboard, go to Security rules → Create rule → Custom rules. Match User Agent contains SeiBot, choose the action Skip, and skip All remaining custom rules, All managed rules, All Super Bot Fight Mode rules, Security Level and Browser Integrity Check. Deploy it, and keep it first in the list so it runs before a rule that challenges.
  • On the Free plan, Bot Fight Mode can't be skipped by any rule. If it is on, turn it off under Security → Settings (filter by Bot traffic), or Sei stays blocked.

Other firewalls and security plugins. Allow requests whose User-Agent contains SeiBot:

  • Sucuri: its firewall allowlists IP addresses and URL paths, not User-Agents, and Sei has no fixed addresses. If it blocks SeiBot, ask Sucuri support to let requests whose User-Agent contains SeiBot through for your site.
  • Wordfence: its rate limiting and its blocking of unknown crawlers can catch SeiBot. Look for SeiBot in Wordfence → Tools → Live Traffic, and loosen the rule that blocked it.
  • SiteGround: its bot defense (the sgcaptcha check) can't be configured per User-Agent by site owners; ask SiteGround support to allow SeiBot for your site.

Sei's servers have no fixed IP addresses, so allow the User-Agent, not an address. Then run Scan now, or Connect WordPress, again.

Contact

Questions or problems with SeiBot: petruswst@gmail.com.