Agent-Ready

Every company has two lists: the Java and Python services that run the business, and the AI agents leadership wants in production this year. The gap between those lists is what this book is about.

An agent is only as useful as the actions it can take, and the actions that matter live inside older services. Agent-Ready shows how to make those services safe for an agent to call: modernize them where they need it, expose them as a small set of governed tools, and prove they behave exactly as they did before.

The method uses IBM Bob as the working partner for reading, planning and changing code, with people deciding what is allowed and verifying what was done. Its one idea: tests protect only what they check.

Inside, you will learn how to:

• Pin current behavior with characterization tests before anything changes
• Find the seams where tools belong in existing code
• Modernize Java without moving behavior
• Wrap services as tools through the Model Context Protocol
• Add guardrails, human approval and audit trails
• Apply the same method to Python services
• Run it as a program across many teams