A document's files — the PDF that arrived, whatever was attached — are listed by their own endpoint, each with a download_url: a signed link that fetches the file directly.
List a document's attachments
From the published contract
curl "https://api.rivet.network/v1/documents/6a5eed00000000000000d001/attachments" \
-H "Authorization: Bearer rk_live_…"| Field | Type | Notes |
|---|---|---|
| id* | string | The attachment's id. |
| filename* | string | The file's name. |
| content_type* | string | null | MIME type, when known. |
| size_bytes* | integer | null | Size in bytes, when known. |
| uploaded_at* | string | null | When it was attached. |
| download_url* | string | A signed, time-limited link to the file. Fetch it promptly; list again when it expires. |
Signed links
download_urlis time-limited — fetch it promptly. When it expires, list the attachments again for a fresh one; the ids are stable.- Don't store or share the URL itself; store the attachment
idand re-list when you need the bytes. - The endpoint follows the same rule as every read: a document your organization isn't a party to answers
not_found, never a hint that it exists.