Module 10 Knowledge Check
5 questions. Pass mark 4 out of 5. Answer every question before checking the answer key below, then retry after reading the feedback.
1. Your capstone must justify each agent by…
- A. Its persona
- B. The distinct capability, tool access, or permission boundary it requires
- C. Its name
- D. Its model
2. The design must state where a human reviews, meaning…
- A. Who reads the logs
- B. The point at which a person can inspect, edit, or refuse before an action occurs
- C. Who owns the code
- D. Who pays
3. Cost controls in the capstone should include…
- A. A cheaper model
- B. Iteration limits, budget caps, and defined behaviour when they are reached
- C. Fewer tools
- D. Shorter tasks
4. Evidence should show…
- A. A successful run only
- B. Per-task outputs, the review point, and at least one handled failure case
- C. The code
- D. The cost
5. The write-up should honestly state…
- A. That it is production ready
- B. What was tested, what was not, and where a human must remain involved
- C. The agent count
- D. The framework version
---
Answer Key and Explanations
Check these only after attempting every question.
1. B - The distinct capability, tool access, or permission boundary it requires
An agent that cannot be justified this way should be merged away.
2. B - The point at which a person can inspect, edit, or refuse before an action occurs
That point is the difference between an assistant and an unsupervised actor.
3. B - Iteration limits, budget caps, and defined behaviour when they are reached
Limits without defined behaviour just convert a cost failure into an unhandled one.
4. B - Per-task outputs, the review point, and at least one handled failure case
How the system behaves when something goes wrong is the more informative demonstration.
5. B - What was tested, what was not, and where a human must remain involved
Naming the untested boundary is what allows someone else to judge readiness.
Practical Check
Apply this module to your own work: complete the module activity for *Capstone: Reviewed Multi-Agent Automation*, then write one sentence naming what your result shows and one naming what it does not.
Strong Answer Pattern
A strong answer names the task, the evidence used, the check performed, and the remaining limitation. It avoids "proved", "guaranteed", or "always" unless the evidence genuinely supports it.