CSV validation usually starts with rules: required fields, types, ranges, allowed
values, and uniqueness.
But if the validation output will be attached to an issue or sent to a teammate,
there is an earlier design question:
Should the report repeat the value that failed?






