Love the tool!
However, with large datasets, it becomes rather time consuming to discover a typo in a column name after several minutes of running, especially if it is towards the end of the config. Could the tool be modified to scan for invalid columns/tables ahead of time? This could save a lot of waiting time.