So if you deploy with a catalog that's not the default (caspersdev), you can't run databricks bundle run cleanup with a parameter.
TODO: get rid of the ephemeral job in the script, create a destroy job as part of the DAB so that you pickup the catalog name in the definition