Summary of the problems I fixed with the task runner:
- I misnamed a command-line flag in some of the test variants
- Near as I can tell, I wasn't supposed to be using resolve() for what I was using it for
- I had a mismatch between the name passed to a template, and the name referenced in a template
- Minor issue about gobbling newlines.
From looking at some of the test failures before I got this all working, I think an important issue is the fact that I'm not confirming my understanding of the terminal colors. If I want proper tests in that area, I'll need to create a command-line flag for controlling the coloring behavior, and make some changes to the way coloring is currently handled.
That seems like a good thing to work on later; I'm already tired right now, so once again I'm not going to push myself.