Section checkpoint

Tools, Function Calling & Agents

12 questions drawn from the 6 chapters in this section. If most land on the first try, you are ready for the next section. If not, the explanation after each answer says which idea to revisit.

Checkpoint

12 questions · first-try answers earn the most XP

1. What happens after the model emits a tool call, before it answers the user?

2. Why does the chapter compare briefing a tool to briefing a clever intern?

3. Why should MCP tool schemas be versioned as they evolve?

4. The chapter says debugging agents without traces is astrology. What is the point?

5. What is a checkpoint for, in an agent system?

6. Your agent books a flight and finds a calendar conflict. What should happen next?

7. What does persisting `{step, tool, args, result}` after every action give you besides crash recovery?

8. Which statement best captures the chapter's three-tier model?

9. An agent retries a create-order tool and produces two orders from one user action. Which single property would have prevented this?

10. Your agent takes 40 tool calls for a task that should need 3. Which failure mode does that most strongly suggest?

11. Why does the chapter say to give each agent the narrowest context it needs?

12. What does peer review between two agents cost, and when is it worth it?

What this checkpoint covers

Next section: Multimodal & Generative AI