Session 1 - Module A: Test Solutions¶
Multiple Choice Test - Module A Solutions
Question 1: What is the key mechanism that prevents infinite loops in reflection agents?
Answer: B) Maximum iteration limits with satisfactory conditions
Explanation: Reflection agents use max_iterations combined with satisfactory conditions (like the "SATISFACTORY" keyword in critiques) to prevent endless self-improvement loops. Without these controls, agents could continuously try to improve their responses indefinitely.
Question 2: In multi-step planning, what is the primary purpose of dependency management?
Answer: B) Ensuring steps execute in the correct order
Explanation: Dependency management ensures that prerequisite steps complete before dependent steps begin execution. The _find_next_executable_step method checks that all dependencies are satisfied before allowing a step to run, maintaining proper execution sequence.
Question 3: What does the planning agent's _parse_plan_text method accomplish?
Answer: B) Converts LLM-generated text into structured PlanStep objects
Explanation: The _parse_plan_text method takes the natural language plan generated by the LLM and parses it into structured PlanStep objects with IDs, descriptions, dependencies, and estimated times, enabling programmatic execution.
Question 4: In multi-agent orchestration, what information is stored in communication patterns?
Answer: B) Summarized message data with sender, recipient, and timestamp
Explanation: Communication patterns store abbreviated message data (truncated to 50 characters) along with metadata like sender, recipient, and timestamp. This allows analysis of interaction patterns without storing full message content, which is maintained separately in message history.
Question 5: What is the three-phase approach used in collaborative task management?
Answer: C) Task analysis, delegation plan creation, plan execution
Explanation: The delegate_complex_task method follows a systematic three-phase approach: first analyzing task requirements to determine needed agent types, then creating a delegation plan for how agents should collaborate, and finally executing the coordinated plan.