The problem with that idea is the implementation: you won't be able to make a game long-term interesting when placing this limits.
(OK, there is a small chance of pulling that off - but I don't think anyone using limited resources has a good chance)
If we ignore all ingame and story-based stuff, there is exactly one reason in the game mechanics why all such games go the route of adding skills (however you call them) to the battlers: to keep the battles interesting and challenging to the player by giving the need to find new uses or combination of skills to fight.
Whenever a new enemy/boss appears, the player has to start looking for a method to win - with different skills and/or elements and similiar effects.
If your game doesn't have skills and only a limited number of options, the player will learn the best fighting tactics sooner - and can keep using them the entire game. To make the battles still interesting instead of a simple button mash, you would have to improve the enemy AI to increase the challenge.
And that is where the impossibility comes in: No one has ever managed to program a good AI, even in chess programs they mostly work with human-created move libraries and throw a lot of processing power into it to check the possible outcomes.
You might be able to keep battles like this interesting if you keep altering the battlefields and the events on them, but that requires a lot of planning and predefining the enemy move AI.
It's simply a fact that many games use changing skills and properties of enemies in order to hide that the enemy AI is usually extremely bad, and if you remove those options from the game you either have to replace them with a different diversion from the AI, improve the AI or have a game that will be no challenge after the first few battles.