switch_model · OpenAI Token Manager MCP: Risk & Policy

WHAT IT DOES

What switch_model does on OpenAI Token Manager MCP

AI agents use switch_model to create or update resources in OpenAI Token Manager MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your OpenAI Token Manager MCP environment.

WHY IT NEEDS A POLICY

Why switch_model needs a policy

The tool changes system state (active model selection) but does not delete data, execute arbitrary code, or move money. It is reversible—users can switch back to a previous tier. While it could have downstream effects on costs and API behavior, the direct action is configuration modification, placing it in the Write category.

From the tool's definition Tool description states it 'Manually switch to the next available model tier', which modifies the configuration or state of which model is being used. This is a reversible configuration change.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

switch_model Rate-limited

switch_model stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "switch_model": {
      "limits": [
        {
          "counter": "switch_model_rate",
          "window": "minute",
          "max": 30,
          "scope": "grant"
        }
      ]
    }
  }
}

LIMIT THIS TOOL →

Instant setup, no code required.

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

FAQ

Questions about switch_model

What does the switch_model tool do? +

Manually switch to the next available model tier. It is categorised as a Write tool in the OpenAI Token Manager MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.

How do I enforce a policy on switch_model? +

Register the OpenAI Token Manager MCP server in PolicyLayer and add a rule for switch_model: 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 OpenAI Token Manager MCP. Nothing to install.

What risk level is switch_model? +

switch_model is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.

Can I rate-limit switch_model? +

Yes. Add a rate_limit block to the switch_model 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 switch_model completely? +

Set action: deny in the PolicyLayer policy for switch_model. 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 switch_model? +

switch_model is provided by the OpenAI Token Manager MCP server (opaldecisionsciences/openai-token-manager-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.