Original practice questions based on this certification's published exam topics — not real exam questions.
11. What is a common cause of an LLM-powered feature producing unexpected output?
12. Why is version-controlling your prompts useful during development?
13. What does "few-shot prompting" involve?
14. What is a reasonable approach when a tool call might fail (e.g., a downstream API is down)?
15. Why might a developer set a maximum number of tool-use steps in an agentic loop?
Claude