AWS Glue job without script, Spark/Scala JAR only

Is there a way to run a Glue job in AWS where all the necessary code is built into a JAR artifact and uploaded to S3?

Right now the best I can do is something like a placeholder wrapper script like

import project.ActualMainClass

object ScriptMain {
    def main(sysArgs: Array[String]): Unit = {
        ActualMainClass.main(sysArgs)
    }
}

It would be nice if I could skip that extra wrapper and provide the JAR and entry point only.