AI Hints
Agentic AI · Lesson 95 of 100

Quiz: Foundations

Agents, tools, and memory.

⏱ 2 min read▶ Video at 2:50:43🎓 Free
Watch this lesson in the Agentic AI courseWatch on YouTube ↗Full Agentic AI course

Time to test yourself. Ten questions on the core of agents. Pause on each one, commit to an answer, and keep score.

Quiz: Foundations
From the video: Quiz: Foundations
  • Ten questions
  • Pause and commit
  • Keep score

Question 1

Question 1. What makes something an agent, rather than a chatbot? A, A longer prompt. B, It takes actions in a loop toward a goal. C, A bigger model. D, A voice. Pause the video and choose. The answer is B. Think, act, observe, repeat.

Quick check
What makes something an agent, rather than a chatbot?

Question 2

Question 2. Who actually runs a tool? A, The model. B, Your code. C, The user. D, The cloud provider. Pause the video and choose. The answer is B. The model only asks.

Quick check
Who actually runs a tool?

Question 3

Question 3. What does the model see of a tool? A, Its source code. B, Its name, description, and inputs. C, Its tests. D, Nothing. Pause the video and choose. The answer is B. That's why descriptions matter so much.

Quick check
What does the model see of a tool?

Question 4

Question 4. Each step works 97% of the time. What happens with twenty steps? A, Still 97%. B, About 54% overall. C, 100%. D, About 90%. Pause the video and choose. The answer is B. Reliability multiplies down.

Quick check
Each step works 97% of the time. What happens with twenty steps?

Question 5

Question 5. How does a model remember earlier messages? A, It stores them. B, Your app resends them. C, It learns them. D, It can't. Pause the video and choose. The answer is B. Models are stateless.

Quick check
How does a model remember earlier messages?

Question 6

Question 6. A tool fails. What should it return? A, Nothing. B, A crash. C, A clear message saying what to do next. D, A guess. Pause the video and choose. The answer is C. Errors are observations.

Quick check
A tool fails. What should it return?

Question 7

Question 7. Why use idempotency keys on actions? A, Speed. B, So a retry can't pay twice. C, Lower cost. D, Better prompts. Pause the video and choose. The answer is B. Repeating is then safe.

Quick check
Why use idempotency keys on actions?

Question 8

Question 8. Which is usually safe to run in parallel? A, Reads. B, Payments. C, Writes to one record. D, Ordered steps. Pause the video and choose. The answer is A. Reads together, writes alone.

Quick check
Which is usually safe to run in parallel?

Question 9

Question 9. Which should you try first? A, A team of agents. B, The simplest workflow that works. C, The largest model. D, A new framework. Pause the video and choose. The answer is B. Use the least autonomy that solves it.

Quick check
Which should you try first?

Question 10

Question 10. What must a sliding window memory risk? A, It costs more. B, It forgets old but important facts. C, It leaks keys. D, It crashes. Pause the video and choose. The answer is B. That's why summaries help.

Quick check
What must a sliding window memory risk?

Summary

Eight or more is excellent. Next, a harder round.

Key takeaways
  • 8 or more: excellent
  • Review what you missed
  • Next: graphs and teams
Next up: Quiz: Graphs and Teams — LangGraph, multi-agent, MCP