AI agents invoke run_python_file to trigger actions in Python Code Runner. 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.
This tool executes arbitrary Python code from a file, which is Execute category. Severity is high because a misused tool could run malicious scripts, modify system state, exfiltrate data, or install malicious packages (especially given pip_install_package is available on the same server). It is not Destructive by itself (depends on what code the file contains), but the execution capability is a severe risk.
From the tool's definition Tool executes Python code ("Run an existing Python file") and the server description confirms it "Enables execution of Python code in a safe environment".
Documented attack patterns abuse exactly the kind of access run_python_file gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Python Code Runner, and nothing reaches the server without passing your rules. This is the rule we recommend for run_python_file:
{
"version": "1",
"default": "deny",
"tools": {
"run_python_file": {
"limits": [
{
"counter": "run_python_file_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}run_python_file 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.
Free to start. No card required.
Run an existing Python file. It is categorised as a Execute tool in the Python Code Runner MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Python Code Runner MCP server in PolicyLayer and add a rule for run_python_file: 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 Python Code Runner. Nothing to install.
run_python_file is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the run_python_file 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.
Set action: deny in the PolicyLayer policy for run_python_file. 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.
run_python_file is provided by the Python Code Runner MCP server (shibing624/mcp-run-python-code). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Python Code Runner, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
4 Python Code Runner tools catalogued and risk-classified — across an index of 43,000+ MCP servers.