Make Terminal Output Your Shared Truth

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-coder-terminal-output-creators

1. When an AI agent's code doesn't work as expected, what is the most effective way to get help?

   1. Share a screenshot of the entire terminal window
   2. Run the code again and hope for a different result
   3. Describe what you think went wrong in your own words
   4. Paste the exact error message and the command that produced it

2. A developer tells an AI agent: 'I want you to write a Python script that fetches data from an API.' Which approach follows the lesson's advice about naming the job before the tool?

   1. Tell the agent to use whatever tool works best
   2. State what the finished script should accomplish (e.g., 'fetch user data from the API and save it to a CSV file')
   3. Ask the agent to use the most popular Python library
   4. Immediately specify which library to use (requests, httpx, etc.)

3. Why does the lesson advise writing the 'smallest useful scope' when working with an AI agent?

   1. Smaller scopes are easier for the agent to complete correctly on the first attempt
   2. The lesson prefers boring, simple projects
   3. Smaller scopes cost less money to run
   4. AI agents cannot handle complex tasks

4. What does it mean to 'run the result as a user, not as a fan of the tool'?

   1. Avoid criticizing the tool's performance
   2. Only use tools made by companies you admire
   3. Support the tool developer's intentions regardless of output
   4. Test the code as if you were a real person using it, not testing whether the tool itself works

5. Before sharing AI-generated code with others, the lesson says you should inspect three things. Which trio is correct?

   1. The syntax, color scheme, and variable names
   2. The diff (changes), data access, and failure path
   3. The cost, the popularity, and the license
   4. The AI's confidence level, the date, and the developer

6. When asking an AI for help with broken code, you should answer four questions beforehand. Which question is NOT listed in the lesson?

   1. What data should the app never expose?
   2. What test proves the change works?
   3. What rollback path exists if the output is wrong?
   4. What programming language should I use?

7. A developer pastes this to an AI: 'build failed.' Why does the lesson consider this unhelpful?

   1. The AI will get confused by too much information
   2. The AI cannot read text that short
   3. Build failures are not important to debug
   4. It removes the specific line numbers and error patterns the AI needs to diagnose the problem

8. What makes terminal output a 'shared truth' between a developer and an AI agent?

   1. It's required by law
   2. It's objective evidence both parties can reference rather than subjective interpretation
   3. It's the only thing the AI can read
   4. Developers must believe whatever the terminal shows

9. The lesson says AI can make a working demo quickly, but 'real skill' is turning that demo into something that is:

   1. Observable, reversible, and safe enough for another person to use
   2. Complex, impressive, and novel
   3. Colorful, interactive, and documented
   4. Fast, cheap, and popular

10. Why is it important to have a 'rollback path' before running AI-generated code?

    1. Because AI always makes mistakes
    2. So you can return to a working state if the new code causes problems
    3. The lesson says AI code is dangerous
    4. Rollback paths are required by IT departments

11. A developer gives an AI agent a massive, complex task all at once. How would the lesson describe this approach?

    1. AI agents prefer complex, detailed tasks
    2. It's the most efficient way to get results
    3. Complex tasks show the developer is serious
    4. It increases the chance the agent will misunderstand requirements and produce errors

12. When debugging with an AI agent, why should you avoid describing what you 'think' happened?

    1. Thinking is not allowed while debugging
    2. AI agents can't understand opinions
    3. The lesson prefers silent debugging
    4. Your interpretation might be wrong and guide the agent toward the wrong solution

13. What does the lesson mean by saying terminal output should be your 'shared truth'?

    1. Terminal output is the only truth in programming
    2. You should only trust output from the terminal
    3. The terminal is always correct about everything
    4. Both you and the AI should reference the same objective output rather than arguing about interpretations

14. If an AI agent produces code that works but you're unsure what it does, what should you do before using it?

    1. Delete it and start over
    2. Inspect the diff, understand the data access, and trace the failure path
    3. Trust that it works since the AI is smart
    4. Use it immediately since testing takes too long

15. The lesson mentions 'observable, reversible, and safe' as criteria for turning an AI demo into something ready for others. What does 'observable' mean in this context?

    1. The code has a graphical interface
    2. You can see what's happening inside the code—logs, outputs, and behavior are visible
    3. The code is open source
    4. The AI shows its work steps