Blog article
Tags, authors, archive pages, and internal graph edges should arrive before the content explosion, not after.
First next step
If this article already matches your situation, move into the tightest internal route instead of jumping straight to a generic contact flow.
Taxonomy is not a cleanup project for later. If blog scale is part of the goal, tags, archive logic, author pages, and internal link clusters should exist before the content count becomes painful.
Even a small seed set benefits from having the same page types that a much larger content program will need. That keeps implementation decisions stable when production volume increases.
Still deciding?
Use the docs layer when you need a clearer decision surface. Use the catalog when you already know the domain. Use diagnostics only when the job needs scoping.
Final CTA
When the content is enough, stay in docs or the scraper catalog. When the job stops being public-safe, take the scoped path.