What the checker looks at
We fetch the page once, as a crawler would, follow up to five redirects and read the first 2 MB of HTML. Then we check the tags that decide how the page is indexed, shown in search, and shared:
- Indexing: HTTP status, redirects, HTTPS, meta robots and the
X-Robots-Tagheader. - The search result: title and meta description, with lengths.
- Duplicates: the canonical URL, and whether it points at this page or somewhere else.
- Structure: H1s, the
langattribute, the mobile viewport and hreflang alternates. - Sharing: Open Graph title, description and image, and the Twitter/X card.
- Machine-readable extras: JSON-LD blocks and their
@types, and whether/llms.txtexists.
The fixes that matter most
Not every warning is equal. In rough order of damage:
- A stray noindex. Staging settings that ship to production are the most common way good pages vanish from Google. If the indexable check fails, fix that first.
- A canonical pointing elsewhere. A template that hard-codes the homepage as canonical tells Google every page is a duplicate of it.
- Missing or duplicated titles. The title is the headline of your search result. Each page needs its own, with the main phrase first.
- Invalid JSON-LD. One trailing comma and the whole block is ignored. Fix syntax before adding more types.
- Relative og:image. Chat apps and social networks need an absolute URL to show a preview image.
What it doesn’t do
It checks one page, not a whole site, and it doesn’t run JavaScript. That second point is deliberate. Google renders JavaScript, but many AI crawlers are widely reported not to, so tags and content that only exist after client-side rendering may be invisible to them. If this checker can’t see a tag, assume some crawlers can’t either, and render it on the server.
Next steps: tune the title and description in the SERP preview, and see which AI bots can reach the page with the AI crawler checker.