code_interpreter · Openai MCP: Risk & Policy

WHAT IT DOES

What code_interpreter does on Openai

AI agents invoke code_interpreter to trigger actions in Openai. 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 code_interpreter needs a policy

A 'code_interpreter' tool almost universally runs arbitrary code in a sandboxed or semi-sandboxed environment. The server context (ChatGPT Plus/Pro bridge with siblings like 'canvas_execute', 'codex_task_create') corroborates this is an execution-class tool. Empty description lowers confidence but the name and server context together make Execute the most likely category.

From the tool's definition Tool name 'code_interpreter' strongly implies execution of code; description is empty so certainty is reduced.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

code_interpreter Rate-limited

code_interpreter 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": {
    "code_interpreter": {
      "limits": [
        {
          "counter": "code_interpreter_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 Openai, so you can run agents against it safely.

FAQ

Questions about code_interpreter

What does the code_interpreter tool do? +

code_interpreter. It is categorised as a Execute tool in the Openai MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on code_interpreter? +

Register the Openai MCP server in PolicyLayer and add a rule for code_interpreter: 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. Nothing to install.

What risk level is code_interpreter? +

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

Can I rate-limit code_interpreter? +

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

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

code_interpreter is provided by the Openai MCP server (robotlearning123/gpt2agent). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.