sandy_run · Sandy MCP: Risk & Policy

WHAT IT DOES

What sandy_run does on Sandy

AI agents invoke sandy_run to trigger actions in Sandy. 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 sandy_run needs a policy

The tool executes arbitrary TypeScript scripts, which is a code execution operation. Although the Sandy server provides sandboxing to mitigate risks (no host credential exposure, cross-account isolation), the tool itself permits running dynamic code whose effects depend entirely on the script arguments. This is a classic Execute category tool.

From the tool's definition Tool description: 'Run a TypeScript script in the Sandy sandbox' — explicitly runs arbitrary code (TypeScript scripts) in an execution environment.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

sandy_run Rate-limited

sandy_run 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": {
    "sandy_run": {
      "limits": [
        {
          "counter": "sandy_run_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 Sandy, so you can run agents against it safely.

FAQ

Questions about sandy_run

What does the sandy_run tool do? +

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

How do I enforce a policy on sandy_run? +

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

What risk level is sandy_run? +

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

Can I rate-limit sandy_run? +

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

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

sandy_run is provided by the Sandy MCP server (jamestelfer/sandy). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.