A multi-agent workflow calls an external pricing API tool that inte…

Question

A multi-agent workflow calls an external pricing API tool that intermittently returns 503s. You need robust tool error handling and fallback so the orchestration does not crash the user session. What should you implement?

Accepted Answer

Detect tool failures, apply retries with backoff where safe, validate tool results before use, fall back to cached prices or a secondary data source, and surface a controlled degradation path to the orchestrator/user when all attempts fail. — Correct answer: A. Tools fail—that’s life. Retries with backoff, result validation, cache/secondary fallback, and a graceful “we can’t price this right now” path keep the orchestrator adult. Hallucinating prices on a 503 is how finance incidents start.

Answer

Ignore HTTP errors and invent prices so the agent always answers confidently.

Answer

Terminate the entire multi-agent application process on the first tool exception.

Answer

Disable error logging to hide intermittent failures from operators.

A multi-agent workflow calls an external pricing API tool that intermittently returns 503s. You need robust tool error handling and fallback so the orchestration does not crash the user session. What should you implement?

Short Explanation

Full Explanation