We should have some tests for the following: - Testing against a snapshot of what we expect the generated shell script to be (using `toMatchSnapshot()`) - Tests which actually run the script against various scenarios