We iterate through each file in a bash for loop or something (since this runs on GitHub's servers), and if anything fails we just fail the workflow,
Or we introduce function that checks the validity of the fed files, and parse each one and check for validity, returning either a pass code or an error code.
There are two ways of doing this. Either;