Hey, sorry for the delayed response! I guess you're right that text games are often written from scratch. Even in my case, I had an idea for a text game I wanted to make before I started writing the engine. But, I thought it would be both beneficial and also fun to structure the code such that it could be used for other games.
Having an idea in mind for a game was really useful for developing the engine, because it gave me a clear picture of the features I would need to implement. You can see this from the very first commit on Github, which included very simple versions of an engine and a game disk to then be expanded upon.
Later, when working on the new features for what would become text-engine 2.0, again I was working in parallel, this time on three new game disks. One of these was what would become the new demo disk for introducing folks to the game, and another was a new story disk I was collaborating on with a friend. Once again, fleshing out those games gave clear direction for which new features to implement next.