Navigation structure checker

The primary navigation region on each crawled page: its entries, their destinations, their accessible names, the sub-menu structure and any entry that leads nowhere.

Consistency across pages: whether the same primary navigation appears everywhere, and whether the key pages are linked from it on every page rather than on some.

Depth: how many clicks from the entry point each crawled page sits at, which is what separates a page that is merely deep from a page that is effectively unreachable.

Orphan candidates: pages the crawl reached that no other crawled page links to.

The footer region and the links it carries — contact, legal, privacy — which is where visitors look before deciding whether to trust a site.

A menu that renders correctly on the home page and loses two entries on article pages is a template condition, and it is invisible to anything that looks at one page.

The same is true of a key page linked from the home page and from nowhere else. Every individual page passes; the site still buries it.

Depth is inherently a whole-site measurement. A page is not deep in itself — it is deep relative to the paths that lead to it.

So the finding is almost always a template or a navigation component rather than a page. Twenty pages reported is usually one component to edit.

One primary navigation, the same on every page, with every entry leading to a real destination.

Every entry with an accessible name — which for an icon-only control means a label a screen reader can announce, since the icon itself announces nothing.

The navigation landmark labelled, so assistive technology can offer it as a region to jump to rather than as an unnamed list of links.

Important pages within about three clicks of the entry point, and breadcrumbs on the deeper sections.

No page reached by the crawl that nothing links to, and a footer carrying the contact, legal and privacy destinations without repeating the same one several times.

Depth and orphan findings are relative to the crawl. A page outside the crawl budget is not measured, and a page linked only from a page that was not reached can be reported as an orphan when it is not one.

Navigation built entirely by script after load may not be present in the HTML that is read, in which case the report describes what was served rather than what a browser eventually shows.

Whether the navigation is actually usable — whether an action is visible, whether a hover menu can be operated — depends on rendered layout, which no static analysis establishes. That is a browser session or a usability test.

An HTML sitemap is reported as present or absent, not as required. Small sites do not need one, and its absence is not treated as a defect.

`structure.menu` analyses the primary navigation region, its links, its ARIA attributes, its sub-menus and its dead entries. `structure.navigation` compares navigation across the crawled pages and checks the key pages are linked from it.

`structure.navigation_ux` measures depth, orphan candidates, breadcrumbs and return-to-home paths from the crawl data already collected.

`structure.footer` detects the footer region and the essential links it carries. `structure.html_sitemap` looks for a human-readable sitemap page, which is a different thing from the XML sitemap the SEO checks audit.

`structure.logo` checks the header brand mark, its alt text, and whether it links back to the home page — the most used navigation control on the web.