Module 09 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. Adding more agents is justified when…
- A. It sounds sophisticated
- B. Distinct roles genuinely need different tools, permissions, or context
- C. The task is long
- D. The prompt is complex
2. In a supervisor pattern, the supervisor's role is to…
- A. Do the work
- B. Route work to the right specialist and decide when the task is complete
- C. Store state
- D. Call tools directly
3. The most common multi-agent failure is…
- A. Too few agents
- B. Agents looping - handing work back and forth without converging
- C. Slow tools
- D. Large state
4. Context passed at handoff should be…
- A. Everything, always
- B. The minimum the receiving agent needs, explicitly structured
- C. Only the last message
- D. The full transcript
5. Permissions in a multi-agent system should be…
- A. Shared by all agents
- B. Scoped per agent so only the one that needs an action can perform it
- C. Held by the supervisor only
- D. Unrestricted internally
---
Answer Key and Explanations
Check these only after attempting every question.
1. B - Distinct roles genuinely need different tools, permissions, or context
Multiple agents add coordination cost. Without genuinely separate concerns, one well-designed graph is better.
2. B - Route work to the right specialist and decide when the task is complete
Separating routing from execution keeps each component's job small and testable.
3. B - Agents looping - handing work back and forth without converging
Handoff cycles need explicit limits and a terminal condition, or they run until the budget dies.
4. B - The minimum the receiving agent needs, explicitly structured
Dumping full history is expensive and buries the actual instruction the next agent must act on.
5. B - Scoped per agent so only the one that needs an action can perform it
Per-agent scoping limits what a confused or manipulated component can actually do.
Practical Check
Apply this module to your own work: complete the module activity for *Multi-Agent Patterns with Supervisor or Handoff*, 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.