I would like to make sure that all of my files are correctly encoded in UTF-8 in a big project repository. Is there a tool for that or a way to do it using unix tools?
In general, there is no way to do this. UTF-8 has no "magic number" or marker, so you can only prove that a file is not in UTF-8 (if it contains invalid sequences), but not that it is.
You can however use a heuristic approach. What exactly works will depend on your data.