Agents And Tool Use easy
An agent has a 'create_refund' tool. A call times out, and the agent loop retries it. What should you add to prevent customers from being refunded twice?
- A An idempotency key per logical action, so the payments API ignores repeated requests.
- B A prompt instruction telling the model never to call create_refund more than once per task.
- C Automatic retries with exponential backoff, since spacing out attempts avoids duplicates.
- D A larger timeout on the tool call so the first attempt always finishes before any retry.
Answer: A. An idempotency key per logical action, so the payments API ignores repeated requests.
A timeout does not mean the refund failed; it may have succeeded, so a blind retry duplicates the side effect. An **idempotency key** lets the downstream API recognize and drop the repeat, at the cost of storing keys and deduplicating server-side. Backoff (C) only changes timing; it is enough for read-only tools, where retrying is genuinely safe.
One system design question a day, with spaced repetition so the ones you miss come back.
Get the appRelated questions
- You are defining a 'schedule_meeting' tool for an agent. Which schema decision most reduces wrong or malformed calls?
- A model emits three parallel tool calls: `get_inventory`, `get_price`, and `get_shipping`. To cut latency, the host appends each result to the transcript as it arrives and immediately re-invokes the model. Runs start failing with malformed-request errors, and some produce hallucinated shipping costs. What should the host do?
- A team splits a coding agent into planner, coder, tester, and reviewer agents, and each receives the full transcript of the previous agent. The expected quality gains did not appear. Cost tripled, and the reviewer often fixates on approaches that earlier agents abandoned. When is a multi-agent split actually justified?
- A support agent should remember each customer's preferences across months of sessions. The team appends every past session summary to the system prompt. After six months, prompts exceed 80,000 tokens, latency has doubled, and the agent sometimes ships to an address the customer replaced weeks ago. What redesign fixes both problems?
- A teammate proposes splitting a single coding agent into five agents, arguing that more agents means smarter results. When is a multi-agent split actually justified?
- A data-analysis agent runs model-generated Python inside the same container as your API service. That container has database admin credentials in environment variables. A security review flags this as critical. What is the best mitigation?