Build your own agent in 30 minutes

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-agentic-ai-ai-build-your-own-agent-30-min-r11a8-teen

1. What is an SDK used for when building an AI agent?

   1. It handles the agent loop, tool calling, and memory automatically
   2. It provides a pre-made user interface
   3. It writes the code for you automatically
   4. It trains a new AI model from scratch

2. What is the recommended approach when adding tools to your first agent?

   1. Define five tools at once to save time
   2. Define one tool, get it working, then add more
   3. Define zero tools since they're optional
   4. Define as many tools as you can think of

3. What happens when an agent needs to perform a task it was designed for?

   1. It shuts down and restarts
   2. It ignores the task and continues
   3. It asks the developer for permission each time
   4. It automatically calls the appropriate tool defined in its configuration

4. Which of these is a key term related to agents that can automatically perform actions?

   1. Video streaming
   2. Data encryption
   3. Cloud storage
   4. Tool calling

5. What is a system prompt in the context of building an agent?

   1. The user interface design
   2. The code that runs on the server
   3. The price charged for using the agent
   4. Instructions that define the agent's role and behavior

6. Why is Vercel AI SDK mentioned as beginner-friendly in agent development?

   1. It works without any coding
   2. It is free forever
   3. It offers a quickstart guide and straightforward setup
   4. It is the only SDK available

7. What should you do after successfully building an agent with one tool?

   1. Push your code to GitHub to share it
   2. Submit it to a competition
   3. Publish it as a paid product immediately
   4. Delete the project and start over

8. What is the main benefit of using an SDK rather than building agent infrastructure from scratch?

   1. It saves time by handling complex loops and memory automatically
   2. It guarantees the agent will be perfect
   3. It eliminates the need for any code
   4. It makes the agent more powerful

9. If you want your agent to perform calculations, what must you create?

   1. A calculator tool that the agent can call
   2. A website to display results
   3. A new AI model trained on math problems
   4. A database of math answers

10. What does it mean when an agent 'calls' a tool?

    1. The agent calls the developer on the phone
    2. The agent sends an email to the tool creator
    3. The agent prints a list of available tools
    4. The agent invokes a function or service to complete a specific task

11. What is the first step mentioned for building an agent in this workflow?

    1. Design a complex user interface
    2. Pick an SDK with a quickstart guide
    3. Train your own language model
    4. Create five different tools

12. What happens if you define too many tools on your first day of building an agent?

    1. The agent will run faster
    2. The tools will automatically work together
    3. You will finish sooner
    4. It becomes harder to debug and verify everything works

13. What is memory in the context of an AI agent?

    1. The storage space for saving files
    2. The ability to retain information across conversations and tasks
    3. The documentation of the agent
    4. The computer's RAM used while running

14. What should you observe after running a prompt that requires a tool?

    1. See if the tool sends you an email
    2. Watch the agent call the tool automatically
    3. Wait for the agent to ask you for help
    4. Check if the internet is still working

15. What is the agent loop that SDKs handle for you?

    1. The circular dependency in your code
    2. The loop between your keyboard and screen
    3. The continuous cycle of receiving input, processing it, and generating output
    4. The file system navigation process