Coding Agents (Like Claude Code) for Real Projects

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-agentic-AI-and-coding-agents

1. What is a coding agent?

   1. A human developer who types very fast
   2. An AI that writes, runs, debugs, and iterates on code autonomously
   3. A spell-checker for code
   4. A tool that only reads code without changing it

2. A coding agent is compared to a 'coding partner who never gets tired.' What does this mean?

   1. The agent will work at all hours without needing breaks
   2. The agent is more creative than human developers
   3. The agent replaces all human coders
   4. The agent is always available for free

3. A developer asks a coding agent: 'Refactor this code to be cleaner.' What does 'refactor' mean?

   1. Delete the code and start over
   2. Rewrite the code to be more readable and maintainable without changing what it does
   3. Add new features
   4. Turn code into a factory

4. Which statement about coding agents is TRUE?

   1. Coding agents never make mistakes
   2. You should always read and understand what the agent changed before using it
   3. Coding agents only work with Python
   4. All coding agent output is production-ready without review

5. What is one benefit of using a coding agent to add a feature to your project?

   1. The feature is always perfect the first time
   2. The agent handles boilerplate and patterns quickly so you can focus on design decisions
   3. You never need to test the feature
   4. The feature is free to build

6. What warning should you keep in mind when using coding agents?

   1. Coding agents always write in the wrong language
   2. Always read the agent's changes — it can introduce subtle bugs or security issues
   3. Coding agents only work on Mac
   4. Never use coding agents on projects that matter

7. What does it mean that a coding agent can 'debug' code?

   1. It removes insects from the computer room
   2. It reads error messages and fixes the code issues causing them
   3. It slows down the program to find problems
   4. It cleans up comments in the code

8. What is the BEST way to review a coding agent's work after it completes a task?

   1. Trust it completely — it's AI
   2. Run 'git diff' to see exactly what changed and review it like a pull request
   3. Check only if tests pass
   4. Only review if the agent asks you to

9. Why should you break large features into smaller chunks when working with a coding agent?

   1. Agents can only write short code
   2. Smaller scopes produce better agent output and are easier to review
   3. Large features are always free
   4. Agents get tired on long tasks

10. What should you provide to a coding agent before it starts a complex task?

    1. Nothing — let it figure everything out
    2. Good context: tech stack, existing code structure, coding conventions, and test framework
    3. Only the feature name and nothing else
    4. A list of other AI tools

11. Which of these is an example of how a coding agent loops to improve its own output?

    1. Write code → submit immediately → never look at it again
    2. Write code → run tests → read failures → fix → run tests again
    3. Write code → delete it → start over every time
    4. Write code → ask the user to test it for them

12. A coding agent writes code that passes all your existing tests. Does that mean the code is definitely correct?

    1. Yes — passing tests means the code is perfect
    2. No — passing tests only means it meets what the tests check, not that it's perfect overall
    3. Yes — AI never writes incorrect code
    4. No — the tests are always wrong when AI writes code

13. Why is treating a coding agent's output like a pull request from a junior developer a good mental model?

    1. Because junior developers are always wrong
    2. Because it reminds you to review carefully — fast and capable but needs human verification
    3. Because coding agents are always slow like juniors
    4. Because it makes you trust the code less than you should

14. What is Claude Code?

    1. A programming language created by Anthropic
    2. Anthropic's coding agent that works in your terminal to write, run, and fix code
    3. A free code editor
    4. A type of database

15. What is the main risk of using a coding agent without reviewing what it changed?

    1. The code will definitely not work
    2. Subtle bugs or security issues could make it into production undetected
    3. The agent will stop working
    4. The code will always be too slow