We should be able to generate a production or preproduction Accounts.json based on information stored in the library registry and information available in the authentication documents of the libraries in question.