run_npm

What run_npm does on Pentester-MCP

AI agents invoke run_npm to trigger actions in Pentester-MCP. 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 run_npm is rated High

npm commands execute code—installing packages, running scripts, or executing arbitrary Node.js. On a penetration testing server designed for autonomous tool execution, this capability poses high risk if an AI agent misuses it (e.g., installing malicious packages, running unvetted scripts). The empty description lowers confidence slightly, but the context and naming strongly indicate code execution.

From the tool's definition Tool named 'run_npm' with empty description on Pentester-MCP server that executes over 200 penetration testing tools. The 'run_' prefix combined with 'npm' indicates execution of Node Package Manager commands, which can run arbitrary code.

The rule that runs run_npm safely

PolicyLayer is an MCP gateway: it sits between your AI agents and Pentester-MCP, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For run_npm, this is the rule to start with:

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

Questions about run_npm