Similarly, the goal of the agent is not defined either. What are we trying to optimize? Maximum points gained ourselves? Minimum points lost ourselves? Maximum point advantage between us and the opponent? Not having our points dropped below a set threshold?

In general, getting the most benefit ourselves and sabotaging the opponents the hardest result in different strategies that cannot be simultaneously optimized. So depending on the goal the optimal strategy is, again, different, so this makes the optimal strategy undefined.

The pay-off matrix of the game is not defined, so the optimal strategy is not well-defined either, hence the kata is unsolvable.

Hi,

Would be nice if you explained exactly how the history is formated in the description!

Needs random tests.