Check manual

For every code you see in a report: what exactly was measured, how strong the evidence behind the recommendation is, and how to fix it.

A check is a single rule the audit verifies on your site. It is not site monitoring; that is repeated tracking of a project over time, and it lives in your dashboard.

For every check we state up front how strong the evidence behind it is:

Necessary condition
Without this, AI search has no way to show you at all. Google states it in its own documentation: a page must be indexed and eligible to be shown with a snippet. The strongest class we have.
Documented effect
There is a measurement with a described methodology that actually found the effect. Not a guarantee of results, but more than an expert guess.
Effect not demonstrated
We recommend it because it does no harm or has some other benefit, but we promise nothing about whether it makes language models cite you. Nobody has demonstrated that yet.

78 checks in total

Crawler access

Structure

Schema.org

Authority

Accessibility

Accessibility findings (A11Y-…)

Codes starting with A11Y- are rules from the axe-core engine, and they have their own continuously updated documentation from its authors; every such finding in the report links straight to it. We deliberately do not give them pages here: ours would be a worse copy that goes stale.

What the axe documentation will not tell you: a finding on one page almost never means one page. The audit walks a sample of your site, and the same defect usually lives in a template: the header, the footer, a product card. Fix the template, not the single page we happened to find it on.

How to read accessibility findings

Recommendations from a language model (LLM-…)

Codes starting with LLM- did not come from a measurement but from a model reading the text of your pages. They cannot have a page per code: their number is merely the position within one audit and may differ next time. One thing about them is worth knowing: they never get critical severity.

How to read recommendations from a language model

The score and how it is calculated

The largest number in the report is a weighted average of five categories, and a quarter of it is a language model's assessment of your content. That makes 100 all but unreachable by design. Where the real ceiling lies, and what to watch instead, has a page of its own.

How the score is calculated and why 100 is not the goal

Audit limits

How many audits you can start per day, why the cap exists and when it renews. It is not a plan restriction: the numbers are the same on every plan and credits have nothing to do with them.

How many audits per day and when the limit renews

Codes from paid modules

SEO, design, colour, AI citations and crawl scope have codes of their own (SEO-…, DSG-…, COLOR-…, CITE-…, SCOPE-…) and each has a manual page too. They only show up in a report when the audit actually ran that module.