What do you guys think?
It's good idea to use CSV format since the main goal is human-readable format and easily processed by average user. But i question the decision where first field (called "Reference" on test vector) may contain multiple types of data. If i'm going to use Excel, i'd rather see additional field which tell type of the data so i could filter/sort it with less effort.
Type,Reference,Label
Transaction,c3bdad6e7dcd7997e16a5b7b7cf4d8f6079820ff2eedd5fcbb2ad088f767b37b,"Withdraw from Binance at 01-01-2021"
Address,1A69TXnEM2ms9fMaY9UuiJ7415X7xZaUSg,"Withdraw address for Binance account"