Thanks szarnyasg. If I've got you here, can you use the ducklake extension commands to get the parquet files for a query without running said query?

That way you could use another query engine while still use duckdb to handle the data mutation.