No dry run

It's hard to debug issues and see what calls are going to be made currently. It would be nice to have a dry run option (potentially called from command line eg. `--dry-run`) to see what would have been executed.
