agent_command · Zion MCP: Risk & Policy

WHAT IT DOES

What agent_command does on Zion

AI agents invoke agent_command to trigger actions in Zion. 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.

PARAMETERS

Parameter	Type	Required	Description
`args`	object	—
`name`	string	Yes

Parameters from the server's own tool schema.

WHY IT NEEDS A POLICY

Why agent_command needs a policy

This tool explicitly allows direct invocation of internal agent commands, acting as an 'escape hatch' that bypasses normal controls. Arbitrary command execution against internal systems represents a critical Execute risk, as the effects are fully dependent on the command arguments and could have wide blast radius including destructive or unauthorized operations.

From the tool's definition "directly invoking an internal agent command" and "Advanced escape hatch"

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

agent_command Rate-limited

agent_command 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_command": {
      "limits": [
        {
          "counter": "agent_command_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 Zion, so you can run agents against it safely.

FAQ

Questions about agent_command

What does the agent_command tool do? +

Advanced escape hatch for directly invoking an internal agent command. It is categorised as a Execute tool in the Zion MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

What parameters does agent_command accept? +

agent_command accepts 2 parameters: args, name. Required: name. The full parameter table on this page comes from the server's own tool schema.

How do I enforce a policy on agent_command? +

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

What risk level is agent_command? +

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

Can I rate-limit agent_command? +

Yes. Add a rate_limit block to the agent_command 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_command completely? +

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

agent_command is provided by the Zion MCP server (zion-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.