I wrote a Python script (it's so shoddy and easy to recreate I won't even share) that will (1) look for definitions as words/expressions between asterisks (2) look for dependencies (words that should be defined beforehand) and (3) produce paragraph-paragraph graph -> minimum spanning tree -> toposort.

The plan is to take five-ish years of blog material that's nearly unreadable to people who haven't followed it from the beginning and produce ordered text.