During design review, operations requires the ability to reconstruct why a multi-agent loan decision path produced a given outcome, correlate spans across orchestrator and tools, and replay problematic runs for debugging. Which observability components should you specify?
Select an answer to reveal the explanation.
Short Explanation
Correct answer: C. Architecture must include observability components: cross-service trace correlation, structured logging of reasoning paths, and replay capture. Host CPU alone (A), printf noise (B), and weekly screenshots (D) cannot explain multi-agent decisions. Specify correlation IDs end-to-end from user request through tools and subagents.
Full Explanation
AI-500 architects specify observability components including cross-service trace correlation, structured logging of agent reasoning paths, and agent replay capture for reproducible debugging. Multi-agent failures often span planner, specialists, MCP tools, and models; without correlation you cannot isolate root cause. Structured events should capture tool names, arguments (redacted), model versions, token counts, and decision branches—balanced against privacy. Replay/capture supports offline debugging and evaluation. Options A, B, and D lack semantic insight into agent behavior. Pair observability with monitoring for coordination health, drift, and automated remediation as separate but related components.