start_game · Zmachine MCP: Risk & Policy

WHAT IT DOES

What start_game does on Zmachine

AI agents invoke start_game to trigger actions in Zmachine. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.

WHY IT NEEDS A POLICY

Why start_game needs a policy

While starting a game appears benign on its surface, it is fundamentally an Execute action because it: (1) instantiates and runs a Z-Machine interpreter process, (2) creates session state that persists across the MCP session, (3) allocates server resources, and (4) could be abused to exhaust resources or trigger unintended game behaviors via malformed input.

From the tool's definition Tool name is 'start_game' and description states it 'Start[s] a new Z-Machine game session', which initializes a game engine instance and creates persistent state (session_id).

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

start_game Rate-limited

start_game stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "start_game": {
      "limits": [
        {
          "counter": "start_game_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches Zmachine, so you can run agents against it safely.

FAQ

Questions about start_game

What does the start_game tool do? +

Start a new Z-Machine game session. Returns the opening text and a session_id. It is categorised as a Execute tool in the Zmachine MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on start_game? +

Register the Zmachine MCP server in PolicyLayer and add a rule for start_game: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Zmachine. Nothing to install.

What risk level is start_game? +

start_game is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit start_game? +

Yes. Add a rate_limit block to the start_game rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.

How do I block start_game completely? +

Set action: deny in the PolicyLayer policy for start_game. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.

What MCP server provides start_game? +

start_game is provided by the Zmachine MCP server (jonathan-meyer/zmachine-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.