Tool Interception

How it works

When an agent requests a tool call, Agent Guard intercepts the call request and evaluates it against your policy before execution. If the tool call is flagged, it can be blocked or escalated without the tool ever running.

Configuration

Enable tool interception in your policy:

{
  "agent_guard": {
    "intercept_tool_calls": true,
    "intercept_tool_results": true,
    "block_on_prompt_injection": true
  }
}