Chain Reaction Issues

[acostarelli]

• Add a testing script to scripts/
• Add support for agent_N_kwargs in instantiating the agents. See [play_game in llm-agents] (https://github.com/Joshuaclymer/GameBench/blob/llm-agents/api/play_game.py) and this commit
• I get the following error:

chain_reaction.py", line 84
    for c in range(cols):
    ^
IndentationError: expected an indented block after 'for' statement on line 83

• Add a blank entry to elo_ratings.json
• I get the following error:

chain_reaction.py", line 76, in get_observation
    rows = len(board) ## default is 11
NameError: name 'board' is not defined

• I get the following error:

random_agent.py", line 10, in take_action
    actions = list(available_actions.predefined.keys())
AttributeError: 'list' object has no attribute 'predefined'

• I get the following error:

chain_reaction.py", line 120, in update
    if (turncount_player1*turncount_player2)>0:
NameError: name 'turncount_player2' is not defined. Did you mean: 'turncount_player0'?

• I get the following error:

chain_reaction.py", line 107, in update
    action = random.choice(list(available_actions.keys()))
AttributeError: 'AvailableActions' object has no attribute 'keys'

• I get the following error:

chain_reaction.py", line 160, in play
    self.turncount_player2 += 1
AttributeError: 'ChainReaction' object has no attribute 'turncount_player2'. Did you mean: 'turncount_player0'?

• When running the game with quick fixes to the above issues + random agents, it appears the game state does not get updated -- it prints the same empty board each turn.