For instance, we currently match on "add/remove" output from git status which is less than fortunate. For performance reasons we don't actually go check if each of those files exists (since sometimes you want to pipe hundreds of results into fpp [EDIT: and are using an expandrive or network mount])