Creating or updating a site no longer fails when its URL has a path

Creating a site, or saving a change to its URL, automatically adds that address to the site's allowed origins: the list that lets the assistant embed on your storefront out of the box. That step only accepts a bare origin (scheme, host, and port), but the site's URL itself is allowed to carry a path, and it did whenever your storefront redirected to a specific page, landing on /en/ or /some/page rather than the root address. That's exactly what happens during onboarding, which probes your storefront's real address for you.

When that happened, the automatic step was rejecting the full address instead of just its origin, so the whole save failed: a new site couldn't be created, an existing one couldn't be updated, and an onboarding submission for a storefront that redirects this way never completed.

That's fixed. Creating or updating a site now always seeds the allowed-origins list with the address's bare origin, whatever path the URL itself points to. Onboarding, manual site creation, and editing a site's URL all work the same regardless of whether your storefront's address includes a path.