Tic-Tac-Toe icon

Tic-Tac-Toe

Enables AI assistants to play tic-tac-toe through a standardized Model Context Protocol interface.

About

This project implements a complete tic-tac-toe game as a Model Context Protocol (MCP) server, offering a standardized interface for AI assistants to interact with the game logic. It allows Large Language Models (LLMs) to seamlessly create, manage, and reset multiple game sessions, make strategic moves, analyze board positions, and retrieve detailed game status. The server exposes a comprehensive set of tools for full game management, supporting AI-human collaboration and integration within MCP-compatible clients like Claude Code.

Key Features

  • Create and manage multiple tic-tac-toe game sessions
  • Supports StdIO, Server-Sent Events (SSE), and Streamable HTTP transport methods
  • Execute strategic moves and analyze board positions via MCP tools
  • Provides tools for game commentary and suggestions for AI assistants
  • 0 GitHub stars
  • Exposes 8 dedicated MCP tools for complete game management (new_game, make_move, get_board, get_status, etc.)

Use Cases

  • Demonstrating AI-human collaboration and MCP integration within LLM environments
  • Integrating a complete game experience into MCP-compatible clients like Claude Code or Claude Desktop
  • Enabling AI assistants to play tic-tac-toe against humans or other AIs