games.tic_tac_toe.generic_engines¶

Generic Tic Tac Toe engines using the new generic player agent system.

This module demonstrates how to use the generic player agent system for Tic Tac Toe, showing the same pattern working across different games with different player identifiers.

Classes¶

TicTacToePromptGenerator

Tic Tac Toe-specific prompt generator using the generic system.

Functions¶

`compare_chess_vs_ttt_patterns`()	Compare the chess vs tic-tac-toe patterns to show generalization.
`create_generic_ttt_config_from_example`(example_name[, ...])	Create Tic Tac Toe engines from predefined examples using generics.
`create_generic_ttt_engines`(player_configs)	Create Tic Tac Toe engines using the generic system.
`create_generic_ttt_engines_simple`([x_model, o_model, ...])	Create Tic Tac Toe engines with simple model configurations using generics.
`create_multi_game_comparison`()	Create engines for multiple games to show the pattern.

Module Contents¶

class games.tic_tac_toe.generic_engines.TicTacToePromptGenerator(players)¶

Bases: haive.games.core.agent.generic_player_agent.GenericPromptGenerator[str, str]

Tic Tac Toe-specific prompt generator using the generic system.

Init .

Parameters:: players (GamePlayerIdentifiers[PlayerType, PlayerType2]) – [TODO: Add description]

create_analysis_prompt(player)¶

Create a Tic Tac Toe analysis prompt for the specified player.

Parameters:: player (str) – Player symbol (“X” or “O”)
Returns:: Prompt template for position analysis
Return type:: ChatPromptTemplate

create_move_prompt(player)¶

Create a Tic Tac Toe move prompt for the specified player.

Parameters:: player (str) – Player symbol (“X” or “O”)
Returns:: Prompt template for move generation
Return type:: ChatPromptTemplate

get_analysis_output_model()¶

Get the structured output model for Tic Tac Toe analysis.

Return type:: type

get_move_output_model()¶

Get the structured output model for Tic Tac Toe moves.

Return type:: type

games.tic_tac_toe.generic_engines.compare_chess_vs_ttt_patterns()¶

Compare the chess vs tic-tac-toe patterns to show generalization.

This function demonstrates how the same generic system works for different games with different player naming conventions.

games.tic_tac_toe.generic_engines.create_generic_ttt_config_from_example(example_name, temperature=0.3)¶

Create Tic Tac Toe engines from predefined examples using generics.

Parameters:

example_name (str) – Name of the example configuration
temperature (float) – Temperature for all engines

Returns:

Dictionary of engines

Return type:

dict[str, AugLLMConfig]

Available examples:

“gpt_vs_claude”: GPT-4 vs Claude
“gpt_only”: GPT-4 for both players
“claude_only”: Claude for both players
“budget”: Cost-effective models
“mixed”: Different provider per role

games.tic_tac_toe.generic_engines.create_generic_ttt_engines(player_configs)¶

Create Tic Tac Toe engines using the generic system.

Parameters:: player_configs (dict[str, haive.games.core.agent.player_agent.PlayerAgentConfig]) – Dictionary of role name to player configuration
Returns:: Dictionary of configured engines
Return type:: dict[str, AugLLMConfig]

Example

>>> configs = {
...     "X_player": PlayerAgentConfig(llm_config="gpt-4"),
...     "O_player": PlayerAgentConfig(llm_config="claude-3-opus"),
...     "X_analyzer": PlayerAgentConfig(llm_config="gpt-4"),
...     "O_analyzer": PlayerAgentConfig(llm_config="claude-3-opus"),
... }
>>> engines = create_generic_ttt_engines(configs)

games.tic_tac_toe.generic_engines.create_generic_ttt_engines_simple(x_model='gpt-4o', o_model='claude-3-5-sonnet-20240620', temperature=0.3, **kwargs)¶

Create Tic Tac Toe engines with simple model configurations using generics.

Parameters:

x_model (str | haive.core.models.llm.LLMConfig) – Model for X player and analyzer
o_model (str | haive.core.models.llm.LLMConfig) – Model for O player and analyzer
temperature (float) – Temperature for player engines
**kwargs – Additional configuration parameters

Returns:

Dictionary of engines

Return type:

dict[str, AugLLMConfig]

Example

>>> engines = create_generic_ttt_engines_simple("gpt-4", "claude-3-opus")
>>> engines = create_generic_ttt_engines_simple(
...     "openai:gpt-4o",
...     "anthropic:claude-3-5-sonnet-20240620",
...     temperature=0.5
... )

games.tic_tac_toe.generic_engines.create_multi_game_comparison()¶

Create engines for multiple games to show the pattern.

This demonstrates how the same configuration approach works across different games with the generic system.