There was the pamphlet in the begging of the game which put keywords into brackets. That could be another way to clarify which elements can be interacted with. So if it's a "[white house] near a lake", I would know right away that trying to swim or draw water from the lake does nothing. If you format the game's text to highlight the keywords in some way, all the reader needs to do is guess the verb.
Just a thought I had in regards to how to keep the same engine but still make the set of actions the player can take a little more clear.