Behavior
Use this endpoint when an app is stuck inerror or provisioning status, or is missing provider infrastructure (e.g. an active fullstack app whose background backend provisioning failed). It checks which provider links are missing and re-provisions them. It is safe to call on a healthy app:
- No hosting project linked → creates the hosting project, seeds
NEXT_PUBLIC_APP_URL, and links it. - Fullstack with no backend linked → starts backend provisioning in the background and returns immediately. This call does not wait for the Supabase project. The
supabaseobject appears onGET /v1/apps/:idonce it is linked. - Nothing to retry →
status: "noop", not an error. Retrying a healthy app is safe and idempotent. - Already-linked infrastructure is left untouched.
provisioning for the duration. It becomes active when this
call finishes, unless the backend is still being provisioned in the background —
in that case the background job sets active on success or error on failure.
Poll GET /v1/apps/:id to find out which.
Response Fields
Errors
501
409
502