sandbox.exec · Dynamic MCP: Risk & Policy

WHAT IT DOES

What sandbox.exec does on Dynamic

AI agents invoke sandbox.exec to trigger actions in Dynamic. 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 sandbox.exec needs a policy

Shell command execution is inherently dangerous because the effects are entirely argument-dependent and can range from benign queries to system compromise, data destruction, lateral movement, or privilege escalation. Even within a sandbox, shell access grants broad capability to an attacker or misbehaving agent.

From the tool's definition Tool description explicitly states 'Run one or more shell commands in an active sandbox session'. The ability to execute arbitrary shell commands is the definition of Execute category.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

sandbox.exec Rate-limited

sandbox.exec 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": {
    "sandbox.exec": {
      "limits": [
        {
          "counter": "sandbox.exec_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 Dynamic, so you can run agents against it safely.

FAQ

Questions about sandbox.exec

What does the sandbox.exec tool do? +

Run one or more shell commands in an active sandbox session. It is categorised as a Execute tool in the Dynamic MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on sandbox.exec? +

Register the Dynamic MCP server in PolicyLayer and add a rule for sandbox.exec: 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 Dynamic. Nothing to install.

What risk level is sandbox.exec? +

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

Can I rate-limit sandbox.exec? +

Yes. Add a rate_limit block to the sandbox.exec 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 sandbox.exec completely? +

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

sandbox.exec is provided by the Dynamic MCP server (mcpland/dynamic-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.