One page can answer at several addresses. The canonical tag says which one is real. This reads the tag on any URL, resolves it, and tells you whether the page is pointing at itself.
We read the raw HTML the server sends, so a tag injected later by JavaScript will not show — and engines treat it the same way.
Reading the page and resolving its canonical…
Nothing to fix here — this part of the page is in order.
A page that names itself is the normal case and settles the question. A page that names a different URL is asking to be dropped from the index in favour of that one — correct for a duplicate, expensive when it happens by accident on a page you want found.
Two canonical tags on one page cancel each other out: engines pick one, or ignore both and decide for themselves. It happens when a theme and a plugin both write one, and nobody sees it in a browser.
Everything the page earns is handed to that domain. On a staging copy, a syndicated article or a migrated site this is sometimes left in by mistake and the page disappears from search with no error anywhere.
We parse the document the server sends, before any script runs — which is what a crawler reads first. A canonical written in by JavaScript is unreliable for exactly that reason.
This is check 1.3 — Canonical tags point to the page itself and check 0.7 — All pages return HTTP 200, run on one page instead of all of them. All 35+ checks are what the agent reads on a site it is installed on, and most of them it can also fix.
On your site, by the agent: what Reflexa Agent does · pricing · install the agent
All of them on one page: free tools.
It tells search engines which address is the real one when the same content answers at several URLs — with and without a trailing slash, with tracking parameters, http and https, www and apex. The address it names is the one that gets indexed and keeps the links.
Every page that can be reached at more than one address benefits from a self-referencing canonical, which in practice means every page. It costs nothing and it stops an engine from choosing a version you did not mean.
A canonical is a hint, not a rule. Engines override it when it contradicts other signals — the sitemap, internal links, redirects or a noindex pointing elsewhere. If it is ignored, those usually disagree with it.
Yes. When a page's canonical is missing or points elsewhere, Reflexa Agent proposes setting it to the page's own address and applies it on your click — a plugin on WordPress, a commit on a coded site.
The free page check reads your public pages from outside and runs all 35+ structural checks — then the agent fixes the ones it holds a hand for, on your click.