A crawler can find a canonical tag and receive a 200 response from the server while links between pages are still broken, according to an article on automating SEO audits. The article describes checking redirects, canonical tags and hreflang as a graph, so unknown targets can be distinguished from errors and evidence can be preserved. It also presents a reproducible Python script with no third-party dependencies, a synthetic example and rules for adding the checks to CI before release.