agent_run · LocalAnt MCP: Risk & Policy

WHAT IT DOES

What agent_run does on LocalAnt

AI agents invoke agent_run to trigger actions in LocalAnt. 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 agent_run needs a policy

This tool delegates control to a local coding agent and can start execution tasks, making it an Execute category risk. The 'execute' mode explicitly triggers automated code execution. While marked deprecated, it remains capable of running arbitrary tasks with side effects.

From the tool's definition Tool description states 'Delegate to a local coding agent' with 'mode=execute starts an execution task.' This directly triggers external code execution whose effects depend on the agent's behavior and the arguments passed.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

agent_run Rate-limited

agent_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": {
    "agent_run": {
      "limits": [
        {
          "counter": "agent_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 LocalAnt, so you can run agents against it safely.

FAQ

Questions about agent_run

What does the agent_run tool do? +

${DEPRECATED} Delegate to a local coding agent. mode=plan returns a plan; mode=execute starts an execution task. It is categorised as a Execute tool in the LocalAnt MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on agent_run? +

Register the LocalAnt MCP server in PolicyLayer and add a rule for agent_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 LocalAnt. Nothing to install.

What risk level is agent_run? +

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

Can I rate-limit agent_run? +

Yes. Add a rate_limit block to the agent_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 agent_run completely? +

Set action: deny in the PolicyLayer policy for agent_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 agent_run? +

agent_run is provided by the LocalAnt MCP server (yuga-hashimoto/localant). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.