Quiz · 11 · Agent Building Blocks
14 questions from 3 pages
These are the Check Yourself questions from each page of the module, collected in course order. Each heading links back to the page the questions test. All module quizzes →
Tool Use & Function Calling
Check yourself
0 / 5 answered
- What does strict mode guarantee about tool arguments?
- In OpenAI strict mode, how do you express an optional parameter?
- Your agent has 180 tools across 12 services and often picks the wrong one. Which change addresses both context cost and selection accuracy?
- Why should a payment tool accept an idempotency key?
- Rewrite this error for the model: {"error": "E_STATE"}
Agent Memory
Check yourself
0 / 5 answered
- A coding agent keeps a list of the steps it has completed in the current task, persisted after each step so it can resume after a crash. Which memory is this?
- The agent learns that this team always wants tests written before code and updates its instructions accordingly. Which kind of long-term memory is that?
- What are Mem0's four memory-update operations?
- Why does Zep store facts with validity intervals instead of overwriting them?
- A user asks your assistant to forget everything about them. What must deletion cover?
Planning & Reasoning
Check yourself
0 / 4 answered
- According to Huang et al. (2024), when does self-correction of reasoning reliably help?
- What is ReWOO's main efficiency gain over ReAct?
- Why keep a plan in harness state when the model already reasons internally?
- Give two re-planning triggers that are better than 're-plan every step'.