run_ssh_command

What run_ssh_command does on SSH MCP Server

AI agents invoke run_ssh_command to trigger actions in SSH MCP Server. 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_ssh_command is rated High

This tool executes arbitrary commands on remote Linux servers via SSH. Execution of remote commands has unbounded side effects depending on the arguments provided—an agent could delete files, modify system configurations, exfiltrate data, or escalate privileges. This is a classic Execute risk and warrants critical severity due to the blast radius on remote infrastructure.

From the tool's definition Tool named 'run_ssh_command' with empty description; sibling tools include 'add_ssh_key', 'issue_cert', 'approve_request', 'cancel_background_job', and 'list_pending_approvals', indicating a privileged SSH management context where executing arbitrary remote…

The rule that runs run_ssh_command safely

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

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

Questions about run_ssh_command