infer

What infer does on OpenClaw

AI agents invoke infer to trigger actions in OpenClaw. 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 infer is rated High

This tool executes arbitrary commands across multiple provider backends (models, media processing, search, embeddings). While not destructive by itself, the ability to run provider-backed commands represents a significant Execute risk — an AI agent could trigger expensive API calls, process sensitive media, or perform searches with side effects depending on what 'model' and 'search' commands are available.

From the tool's definition Tool description states 'Run provider-backed model, media, search, and embedding commands' — the verb 'Run' combined with 'commands' indicates execution of external operations whose effects depend on arguments.

The rule that runs infer safely

PolicyLayer is an MCP gateway: it sits between your AI agents and OpenClaw, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For infer, this is the rule to start with:

{
  "version": "1",
  "default": "deny",
  "tools": {
    "infer": {
      "limits": [
        {
          "counter": "infer_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

Questions about infer