Data quality crawler
Have a script that crawls over new commits in data folders and looks for inconstancies in the data. Basically looking for csv files to see if they contain errors that would prevent standard python tools from importing them. This is essentially data quality checker that flags errors for NaN values or badly formated files.
Might be a bit hard to do in a general way. One workaround is to search for common csv file types while excluding folders that have a skip tag 'skip_dq' file touched into the dir.
This may already have a package or elements written by someone else.