I really wish it could infer the language from the file type. One of the things that prevents me from reaching for semgrep as often as I want to is the complexity, and having it infer language from filetype would be nice.