"Give these a genuine try before checking the answers — recalling an idea locks it in far better than re-reading it."
Learning Objectives
By the end of this module, you will be able to:
- Test your understanding of the whole track
- Spot any topics worth revisiting
- Reinforce key ideas through active recall
How to Use This Quiz
Answer all 10 questions first, then scroll to Answers & Explanations. Each miss points you to the exact module to review.
There's a reason quizzes beat re-reading. Pulling an answer from your own head — even a wrong one — forces your brain to reconstruct the idea, and that effort is what makes it stick (psychologists call it the "testing effect"). So resist the urge to peek; a genuine attempt teaches more than a confident skim.
Try this: Decide your answer before looking. If you're unsure between two options, note why — that hesitation shows you exactly what to revisit.
The Questions
1. What best describes an AI agent?
- A. A faster search engine
- B. Software that uses an LLM to plan and take actions toward a goal
- C. A database of answers
- D. A spellchecker
2. The biggest difference between an agent and a chatbot is that an agent…
- A. Is always correct
- B. Only works offline
- C. Takes actions over multiple steps, not just replies
- D. Cannot use language
3. The "agent loop" is best summarized as…
- A. Think → Act → Observe → repeat
- B. Download → Install → Run
- C. Ask → Wait → Forget
- D. Train → Test → Deploy
4. What do "tools" let an agent do?
- A. Nothing useful
- B. Take actions in the outside world (search, email, code…)
- C. Only change its font
- D. Slow itself down
5. In function calling, who actually runs the tool?
- A. The LLM runs it internally
- B. Your app runs it and returns the result to the model
- C. Nobody — it's imaginary
- D. The user runs it by hand
6. What does RAG give an agent?
- A. A faster processor
- B. Relevant facts to answer from, like an open-book exam
- C. A new personality
- D. Permission to skip prompts
7. Short-term context differs from long-term memory because context…
- A. Lasts forever
- B. Holds the current task; memory persists across sessions
- C. Is only for images
- D. Cannot be used by agents
8. Prompt injection is…
- A. A way to speed up the model
- B. Hidden malicious instructions in content the agent reads
- C. A type of database
- D. A helpful feature
9. Which is a good guardrail for a risky action (like spending money)?
- A. Let the agent decide alone
- B. Require human approval before it acts
- C. Give it maximum access
- D. Turn off all logging
10. When is a simple workflow better than an autonomous agent?
- A. When the task is the same predictable steps every time
- B. Never
- C. Only for images
- D. When you want it to be slower
Answers & Explanations
- B — An agent uses an LLM to plan and act toward a goal. (Module 1)
- C — Agents take actions across multiple steps; chatbots just reply. (Module 1)
- A — Think → Act → Observe → repeat, until the goal is done. (Module 2)
- B — Tools are the agent's actions in the real world. (Module 5)
- B — The LLM requests a tool; your app runs it and returns the result. (Module 5)
- B — RAG supplies relevant facts to answer from — an open-book exam. (Module 7)
- B — Context = the current task's working memory; long-term memory persists. (Module 6)
- B — Prompt injection hides malicious instructions in content the agent reads. (Module 9)
- B — Require human approval for high-risk actions. (Module 9)
- A — For predictable, repeatable steps, a fixed workflow beats an autonomous agent. (Modules 4 & 8)
How did you do?
- 8–10: Excellent — you understand how agents think and act. You're ready for the projects in Module 10.
- 5–7: Good foundation — revisit the modules behind any misses, then retry those from memory.
- 0–4: Skim Modules 1, 2, and 5 again, then retry. It'll click.
Notice how the questions cluster: 1–3 test what an agent is and how it loops, 4–5 cover tools, 6–7 cover knowledge and memory, and 8–10 cover safety and the right approach. If your misses bunch in one cluster, you've found the topic most worth another pass.
One more habit worth building: check whether you got each answer right for the right reason. It's easy to pick the correct option by eliminating the silly ones without really understanding the idea. For each right answer, explain to yourself why the wrong options are wrong too. If you can, the concept is genuinely yours; if not, that's a quiet signal to revisit the module.
Key Takeaway: If you can answer these, you understand what an agent is, the loop it runs, how tools and function calling work, what RAG and memory add, and how to keep agents safe. Any misses point straight to the module worth another look.