for a number of databases we have a lot of post-deployment steps - basically to add conditional data to tables (that cant be solved by table linking) or adding SQL Jobs.
Currently this can all be thrown in a post-deploy script, but this makes it monolithic and much harder to see changes in objects.

SQLCMD syntax would solve this problem (:r .filename.sql), but even executing everything in the post-deploy directory in alpha order would make the post-deploy much more manageble