Whenever an internal table is created for the core clinvar entity types, go ahead and create a snapshot. We only pay for the difference in data between the table and the snapshot, so free.
No point in creating a snapshot of the processing_history table upon creation because it is intended to change. We could potentially store a like a rolling window of 4 weekly snapshots of processing_history just in case something catastrophic happens we'll have a few past versions.
https://cloud.google.com/bigquery/docs/table-snapshots-create