hosting
POST /v1/sites/{siteId}/backups/{backupId}/download
Mint a time-limited download URL for one of a site's backups.
Authentication
Send an API key as a bearer token. The key must carry the hosting.backup.export permission; a key without it is refused with 403, not 404.
This endpoint takes no organisation id. Your key already identifies the organisation it belongs to, and the response is scoped to it.
Try it
Replace anything in angle brackets with your own values, and the key placeholder with a key from your dashboard.
curl -X POST https://api.zinndigital.com/v1/sites/{siteId}/backups/{backupId}/download \
-H "Authorization: Bearer zdk_live_…"Signed in? The API console in your dashboard fills in your real organisation id and your own key, and runs the request against the live API so you can see the actual response. Open this endpoint in the API console
Details
Returns a presigned URL the customer's browser fetches **directly from object storage**; the archive never passes through the control plane. A site archive is measured in gigabytes, so proxying it would make the customer's site size our memory ceiling and hold a request thread for the length of the transfer. ⛔ **`POST`, despite reading nothing.** The response body is a bearer credential for the whole site — database included — so a `GET` would invite it into browser history, proxy logs and shared caches, and the audit row this writes would be created by every prefetch. The grant expires; take a new one rather than storing it. A backup that never produced a stored archive is `422` (`backup_not_downloadable`) rather than a signature over a key that is not there, which object storage would reject seconds later as an opaque `404`. Requires `hosting.backup.export` — or `hosting.backup.manage`, which every holder of the old gate still has — and never `sites.view`: reading a tenant's database out is never a viewer's. A per-site *editor* holds `hosting.backup.export`, which is how `zinnector pull` brings the files down for local development without being able to restore.
Parameters
| Name | Type | Required | What it is |
|---|---|---|---|
siteId (path) | Uuid | Yes | Site ID (UUIDv7). |
backupId (path) | Uuid | Yes | The backup's id (UUIDv7), as `getSiteBackups` reports it. **Ours**, minted when the row was written — never the storage key, which is an internal object path and is deliberately… |
Response
| Name | Type | Required | What it is |
|---|---|---|---|
url | string | Yes | The presigned object-storage URL to fetch the archive from. Hand it straight to the customer's browser and let it expire; take a fresh grant rather than caching this one. |
expires_at | string | Yes | When the signature stops working. A download that has already started is unaffected; a new one after this moment is refused by storage. |
size_bytes | integer | Yes | The archive's size, as object storage reported it when the backup was stored — so a client can warn before a multi-gigabyte download. |
etag | string | Yes | The stored object's entity tag, for a client that wants to verify the bytes it received are the bytes we hold. Empty when the backend did not report one. |
Errors this endpoint can return
401 · 403 · 404 · 422 · 429