MCTS MCP Server

Instructions

Implementation Reference

• src/mcts_mcp_server/server.py:172-236 (handler)

  The primary handler function that executes the initialize_mcts tool logic. Validates the input parameters, checks for required Gemini API key, initializes and updates the global server_state dictionary with the provided question, chat_id, provider, and model, resets MCTS progress counters, sets initialized flag to True, logs the initialization, and returns a status dictionary with configuration details.

  async def initialize_mcts(
      question: str,
      chat_id: str = "default",
      provider: str = "gemini",
      model: str | None = None
  ) -> dict[str, Any]:
      """
      Initialize MCTS for a question.
  
      Args:
          question: The question or topic to analyze
          chat_id: Unique identifier for this conversation session
          provider: LLM provider to use (currently only 'gemini' supported)
          model: Specific model name to use (optional, defaults to gemini-2.0-flash-lite)
  
      Returns:
          Dict containing initialization status, configuration, and any error messages
  
      Raises:
          Exception: If initialization fails due to missing API key or other errors
      """
      try:
          # Validate inputs
          if not question.strip():
              return {"error": "Question cannot be empty", "status": "error"}
  
          if provider.lower() != "gemini":
              return {"error": "Only 'gemini' provider is currently supported", "status": "error"}
  
          # Check if API key is available
          api_key = os.getenv("GEMINI_API_KEY") or os.getenv("GOOGLE_API_KEY")
          if not api_key:
              return {
                  "error": "GEMINI_API_KEY or GOOGLE_API_KEY environment variable required",
                  "status": "error",
                  "setup_help": "Set your API key with: export GEMINI_API_KEY='your-key-here'"
              }
  
          # Update state
          server_state.update({
              "current_question": question,
              "chat_id": chat_id,
              "provider": provider.lower(),
              "model": model or "gemini-2.0-flash-lite",
              "iterations_completed": 0,
              "best_score": 0.0,
              "best_analysis": "",
              "initialized": True
          })
  
          logger.info(f"Initialized MCTS for question: {question[:50]}...")
  
          return {
              "status": "initialized",
              "question": question,
              "chat_id": chat_id,
              "provider": server_state["provider"],
              "model": server_state["model"],
              "message": "MCTS initialized successfully. Use run_mcts_search to begin analysis."
          }
  
      except Exception as e:
          logger.error(f"Error initializing MCTS: {e}")
          return {"error": f"Initialization failed: {e!s}", "status": "error"}

• src/mcts_mcp_server/server.py:85-98 (registration)

  MCP tool registration within the list_tools handler. Defines the tool name, description, and input schema specifying required 'question' parameter and optional 'chat_id', 'provider', 'model' fields.

  types.Tool(
      name="initialize_mcts",
      description="Initialize MCTS for a question",
      inputSchema={
          "type": "object",
          "properties": {
              "question": {"type": "string", "description": "The question to analyze"},
              "chat_id": {"type": "string", "description": "Unique identifier for this conversation", "default": "default"},
              "provider": {"type": "string", "description": "LLM provider", "default": "gemini"},
              "model": {"type": "string", "description": "Model name (optional)"}
          },
          "required": ["question"]
      }
  ),

• src/mcts_mcp_server/server.py:88-97 (schema)

  Input schema definition for the initialize_mcts tool, outlining the expected JSON object structure with properties and requirements.

  inputSchema={
      "type": "object",
      "properties": {
          "question": {"type": "string", "description": "The question to analyze"},
          "chat_id": {"type": "string", "description": "Unique identifier for this conversation", "default": "default"},
          "provider": {"type": "string", "description": "LLM provider", "default": "gemini"},
          "model": {"type": "string", "description": "Model name (optional)"}
      },
      "required": ["question"]
  }