claude_prompt · Claudecode MCP: Risk & Policy

WHAT IT DOES

What claude_prompt does on Claudecode

AI agents invoke claude_prompt to trigger actions in Claudecode. 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 claude_prompt needs a policy

This tool executes external commands (Claude Code CLI operations) based on user-supplied prompts. While the tool itself may not directly modify filesystems or run arbitrary shell commands, it delegates execution to an external CLI tool whose effects depend entirely on the prompt arguments provided.

From the tool's definition Tool description states it 'Run[s] a one-shot prompt against the Claude Code CLI in headless' mode. The server wrapper enables users to 'execute Claude Code commands' with 'prompt-based interactions.' The tool invokes an external CLI (Claude Code) to process…

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

claude_prompt Rate-limited

claude_prompt 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": {
    "claude_prompt": {
      "limits": [
        {
          "counter": "claude_prompt_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 Claudecode, so you can run agents against it safely.

FAQ

Questions about claude_prompt

What does the claude_prompt tool do? +

Run a one-shot prompt against the Claude Code CLI in headless,. It is categorised as a Execute tool in the Claudecode MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on claude_prompt? +

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

What risk level is claude_prompt? +

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

Can I rate-limit claude_prompt? +

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

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

claude_prompt is provided by the Claudecode MCP server (trevoraspencer/claudecode-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.