Quickstart

What you are installing#

A single generated <script> tag, scoped to one domain by a capability token. It carries no inline code — the tag points straight at the module, which the browser loads asynchronously, out of the way of your page. See the embed script for exactly what that module does.

1. Copy your line#

Every domain on your account has its own line, generated from your install page once signed in. It is not interchangeable between domains — a line copied for one hostname does nothing on another. The install page asks where your site lives first, then shows the one artifact for that answer.

2. Paste it in your page head#

Paste it inside <head>, on the template every page shares — a theme header, a root layout, or a tag-manager container. It needs to run once per page, so the placement is whatever your platform already uses for site-wide scripts:

• WordPress — a header plugin (WPCode) or Appearance → Theme File Editor on a child theme, before </head>.
• Shopify — Online Store → Themes → Edit code → theme.liquid, before </head>.
• Webflow or Wix — the project’s custom-code setting for the page head, on every page.
• Squarespace — Settings → Advanced → Code injection, in the Header box.
• Your own code — the HTML template every page shares, inside <head>.
• An AI coding agent — the install page generates a ready prompt that carries the line and the placement rules; paste it into Claude Code, Cursor or similar.

3. Confirm it is running#

Presence in your HTML is not the finish line — the line has to actually run in a browser. Check my install on your install page runs a live check: it fetches your homepage server-side, looks for the line in the raw HTML, and cross-references whether the module has recently called home from a real page view — and, when the page says so, from which host. We also run that same check about once a day on our own. The install screen re-reads the stored result every few seconds while you have it open — that polling is not a fresh look at your site, only a refresh of what was last recorded; open any page of your own site in another tab and it turns over within a few seconds of the module running there.

Your install page names one of five states:

• Running — the line is live and we are hearing from it. This is the only state that means it is working.
• Running elsewhere — the line is running, but on a host that is not your site. Usually a staging or development copy. It works; it has not reached the live site yet.
• Found, not running — the line is in your HTML but has not run in a browser yet. Usually a content security policy: it needs to allow us in script-src https://seogence.com and connect-src https://seogence.com. An ad blocker or privacy extension in your own browser can do the same — the line still runs for everyone else.
• Not found yet — we have not seen the line on your homepage yet.
• Could not check — we could not reach your site to look. That is about our request, not your install; behind a firewall, a bot filter or a robots.txt that keeps us out, the line may already be running.

One limit worth knowing: the fetch reads your homepage, not every page, so an HTML-only miss tells you about your homepage specifically. An install our fetch cannot see at all — injected through a tag manager, or added by a single-page app — still confirms on its own as soon as one real page view runs it.

We respect your robots.txt. If it disallows SeogenceInstallVerifier, the check cannot read your homepage, and your install screen says so. To let us in, add User-agent: Seogence with Allow: /. Rules that name SeogenceInstallVerifier or SeogenceBot come first, so if you have any, allow / there too. Once the check has seen the line pasted in your own HTML, page captures read your pages regardless of robots.txt: you invited us. Until then, and for a line added through a tag manager, they respect it too.

What happens next#

A page enters the system when a visitor opens it and the capture runs — there is no scan or crawl that walks your site in the background. Three conditions apply:

• Pages carrying noindex, or sitting under a private path prefix such as /cart, /checkout, /account or /login, are skipped by design.
• Capture waits for the page to load and settle, so a visitor who leaves within about two seconds does not register the page.
• Each account has a page quota; once it is reached, further pages are not added.

An admin-triggered sitemap import can seed a larger set of pages at once; ask support if you want that run for your domain.