Crawld and Next.js
Fixes as a pull request, built before you see it.
Status: Connector in progress . Today, scan a Next.js site free with no access granted, and every fix arrives written out for you to apply in your codebase. Below is how the connector will deliver changes once it ships, and what it can honestly verify.
How fixes will arrive
A branch, a pull request, and the result of a real build.
What the rubric can see here
- Metadata from
generateMetadataand static objects - Canonicals, Open Graph, schema from layouts
- App and Pages Router
- Whether a route ships content in its HTML
Limits worth knowing
- Content from a headless CMS is a Mode D problem inside a Mode A site
- A fix that depends on runtime data cannot be verified by a build alone
Where this sits
| Mode | What we get | How changes arrive | Verified | Status |
|---|---|---|---|---|
| A · Repo write | A repo we can push a branch to | Pull request; you review and merge | Build | In progress |
| B · Read-only repo | Read access plus somewhere to push | Branch or fork you own | Build | In progress |
| C · Cloud mirror | A snapshot or export | Patch or PR from a clone | Build (mirror) | In progress |
| D · CMS drafts | OAuth into a hosted CMS | Unpublished drafts; you publish | Preview | In progress |
| E · URL only | A public URL | Written-out fixes; you apply | Advisory | Live |
Build-verified means a build ran. It applies to A, B and C and nothing else. A hosted CMS has no build, so D is preview-verified and E is advisory. Those labels are never blurred.
What you can do today
Scan the site. The free scan needs no access to Next.js. It reads the published HTML the way a search or answer engine does, runs all 129 checks, and reports what it could not measure rather than counting it as a pass. Every finding names the page and the change, so it is actionable by hand whether or not a connector ever exists.