Tangential: Parsing the transcribed text would be a way to add programming actions to the text. Maybe /bin/audio-sh? My favourite parsing technology, ATM, is Ohm-JS.
Descript.com sells an audio/video editing/mixing-board tool based on textual editing (like, say, Logic, GarageBand, iMovie, Da Vinci Resolve, etc., but text-document-based). One could write a quickie parser that transpiles specific text phrases to /bin/bash scripts (or Python, or, ...).