Looking for advice on a metadata discrepancy in bulk export jobs for MMS attachments. The /api/v2/recordings/bulkexport endpoint returns a 202 Accepted, but the resulting S3 manifest files lack the original_content_type header for image attachments. This breaks our chain of custody verification script. The export job completes successfully with no errors in the audit trail. Any ideas on why the metadata stripping occurs during the ingestion phase?