chat_completion · Litellm MCP: Risk & Policy

WHAT IT DOES

What chat_completion does on Litellm

AI agents invoke chat_completion to trigger actions in Litellm. 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 chat_completion needs a policy

This tool executes inference requests against LLM models, triggering external API calls with potentially arbitrary prompt content. While it doesn't directly delete data or move money, it can be misused to exfiltrate data, generate harmful content, or chain with other tools (e.g., key_generate).

From the tool's definition "Generate chat completions using LiteLLM. Supports all models available in your LiteLLM instance."

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

chat_completion Rate-limited

chat_completion 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": {
    "chat_completion": {
      "limits": [
        {
          "counter": "chat_completion_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 Litellm, so you can run agents against it safely.

FAQ

Questions about chat_completion

What does the chat_completion tool do? +

Generate chat completions using LiteLLM. Supports all models available in your LiteLLM instance. It is categorised as a Execute tool in the Litellm MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on chat_completion? +

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

What risk level is chat_completion? +

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

Can I rate-limit chat_completion? +

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

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

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