How a site with no CMS still gets a blog

This page exists to prove one thing end to end: an article published by the platform onto a site that has no CMS, no SDK and no build step, served through a single rule at that site's own edge.

What had to be true

The route had to be created by a surface that only ever answers for URLs it created. The head tags on this page had to come from the same delivery ledger every other payload goes through. And the bytes had to be in the raw HTML, not assembled by JavaScript, because the crawlers that matter run none.

If you are reading this sentence in view-source, all three held.