StoredDocument.source records where a document came from. The render endpoint sets it from options.category (defaulting to unknown, or batch for batch renders), so an integrating product can tag every PDF it generates with a stable label. You filter on it with GET /api/documents?source=... and target it for bulk cleanup with DELETE /api/documents?source=.... Because source is indexed both alone and paired with tenantId, querying a tenant's documents by source stays cheap even as the table grows.
Skarn PDF
StoredDocument.source
The originating label for a StoredDocument — used to filter and bulk-manage documents.