What this check inspects
Shipwork reads one page, extracts every link, skips mailto, tel, javascript and same-page anchors, removes duplicates, and requests the first 25 unique URLs. It reports links that return an error or do not respond at all, and separately links that still redirect instead of answering directly.
What a failure means
A link that returns 404 or above, or fails to connect, is an error. It sends visitors to a dead end and gives the crawler a page it cannot use. A link that redirects is a warning: it works, but it costs a hop and usually means the destination moved without the link being updated.
How to fix it
- Update the link to the current URL, or remove it if the destination is gone.
- Point redirecting links straight at the final URL so they do not chain.
- Check the links after a catalogue or blog migration, when handles and slugs often change.
- Prefer fixing the source over adding a redirect, if the same link is generated from a template.
A typical failure, worked through
The setupA store renames a collection from /collections/summer-sale to /collections/sale and adds a redirect, but the footer still links to the old path.
What the check reportsThe check reports the footer link as a redirect rather than broken, and if the redirect is later removed it becomes a hard 404. Either way the link is not answering directly.
The pointRedirects hide link rot until they are cleaned up. Point links at the final URL so removing the redirect does not break the page.