Defensive Tool Design

The key to building better AI agents.

A practical field guide for putting safety into the tool boundary before the agent ever gets near production authority.

The book shows how to put safety into the tool itself: purpose, schema, permissions, side effects, approvals, retries, audit trails, observability, and evals. Prompts can ask for care. Boundaries make care enforceable.

Buy KindleBuy PaperbackGet the Agent Skill

About the author:Blair Hudson is an Australian AI and software engineering leader working on generative AI at one of Australia's largest banks. His work sits where agents, platform engineering, software architecture, security, and regulated production systems meet.