Agents That Write Code

End-of-lesson check

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

1. What distinguishes coding agents like Cursor from simpler AI code helpers?

   1. They can write and test code across multiple files automatically
   2. They require a human to run every command manually
   3. They work without any internet connection
   4. They only suggest code for Python programs

2. A developer tells a coding agent 'add a search feature.' What typically happens next?

   1. The agent immediately submits the code to the live website
   2. The agent writes code in several files, runs tests, and fixes any failures
   3. The agent deletes existing code to start fresh
   4. The agent asks the developer to write all the code first

3. Which of these is NOT one of the three main capabilities of coding agents mentioned in the material?

   1. Writing documentation for user manuals
   2. Suggesting refactors based on best practices
   3. Writing code across multiple files
   4. Running tests and fixing failures

4. What role do humans play in the workflow with coding agents?

   1. They review what the AI drafts before changes are finalized
   2. They write all the code while the AI watches
   3. They only fix mistakes the agent cannot find
   4. They are no longer needed for software development

5. Which of these is an example of a coding agent?

   1. Microsoft Word
   2. Netflix
   3. Slack
   4. Claude Code

6. What does the term 'PR' stand for in the context of coding agents?

   1. Python Repository
   2. Program Repair
   3. Public Release
   4. Pull Request

7. Why might a company choose to use coding agents?

   1. Because they work without any electricity
   2. Because they can replace all human programmers
   3. To speed up the process of writing and testing code
   4. Because they cost more than human developers

8. When a developer gives a coding agent a goal like 'add a login button,' how many files might the agent edit?

   1. Only one file at most
   2. Exactly one file per feature
   3. Zero files, since it only suggests ideas
   4. It might edit several files to complete the feature

9. What happens during 'review' in the coding agent workflow?

   1. A human checks the AI-generated code before it gets merged
   2. The tests review each other for errors
   3. The AI reviews human-written code
   4. The code is automatically published without checking

10. What makes coding agents different from basic autocomplete tools?

    1. They only work on weekends
    2. They can run tests and fix bugs they discover
    3. They cannot handle programming languages
    4. They require more typing from developers

11. If a coding agent's tests all pass, what is the typical next step?

    1. The agent automatically launches the website
    2. The code is ready for human review
    3. The agent publishes the code immediately
    4. The agent shuts down until new instructions

12. Which statement best describes how software gets built with coding agents?

    1. AI and humans never communicate
    2. Humans review what AI drafts
    3. AI writes code and publishes it without any oversight
    4. Humans write all code and AI only checks spelling

13. What type of task would be LEAST suited for a coding agent?

    1. Adding a new feature to an app
    2. Fixing a bug in existing code
    3. Writing a novel about emotions
    4. Running automated tests

14. What is one benefit of having a coding agent suggest refactors?

    1. It makes code run slower on purpose
    2. It deletes unused variables randomly
    3. It converts code to a different programming language
    4. It can identify improvements humans might miss

15. Based on the material, what is the biggest change coding agents bring to software development?

    1. Software development stops using tests
    2. Code is written entirely by computers without humans
    3. All programmers lose their jobs
    4. Humans shift from writing code to reviewing AI-drafted code