run

What run does on Github

AI agents invoke run to trigger actions in Github. 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 is rated High

The 'run' tool executes scripts or files, which is inherently an Execute action. The ability to run arbitrary scripts presents significant risk if an AI agent is manipulated into executing malicious code, accessing credentials, or modifying repository state. Severity is high due to the broad potential for unintended side effects depending on script arguments and environment.

From the tool's definition Tool name 'run' with description 'Runs a script or file with' indicates execution capability. On a GitHub MCP server context with siblings like 'ansible-playbook' and 'api', this performs arbitrary code execution.

The rule that runs run safely

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

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

Questions about run