Agent Exercise

Question 1

What was the goal of the Base64 exercise?

Answer 1

To control the LLM’s output format by instructing it to respond only in Base64, not natural language.

Question 2

What is Base64?

Answer 2

A text encoding method that converts data into ASCII characters for safe storage or transmission.

Question 3

Why does Base64 matter for Agentic AI?

Answer 3

It trains you to enforce strict format rules for LLM outputs — critical for structured actions in agent loops.

Question 4

What message role should set the formatting rules?

Answer 4

The “system” message.

Question 5

What message role gives the instruction or question?

Answer 5

The “user” message.

Question 6

What library function sends prompts to the LLM?

Answer 6

completion() from the LiteLLM library.

Question 7

How can you verify that a response is Base64?

Answer 7

Decode it using Python’s base64.b64decode() function.

Question 8

What does decoding the Base64 string reveal?

Answer 8

The original English message encoded by the AI.

Question 9

Why is this exercise important?

Answer 9

It builds confidence in controlling AI behavior using clear, rule-based prompting.

Question 10

What concept does this practice connect to in Agentic AI?

Answer 10

Reliable communication between reasoning (LLM) and execution (Python) layers.