Chapter 61 — Secure AI Agents & Tool Execution

61.1 Introduction Traditional AI inference usually follows: User ↓ Prompt ↓ Model ↓ Response An AI agent introduces additional capabilities: User ↓ Agent ├── Reasoning ├── Memory ├── Retrieval