Conduit currently provides an API for processing Remarkup, but it only lets you process one block of text. Thus for each set of Remarkup you want to process, you have to call the API.
There should be a `remarkup.processbulk` API to process large sets of Remarkup from external applications.