Currently a prefix for the identifier column can be specified in the config with recordIDPrefix": "kbr|".
Yet this prefix is only appended to the main CSV output, not to the identifier column of the other output CSV files. This makes the output inconsistent and makes later joins more difficult.