Q15 of 60 — CCA Practice Exam

Multi-Agent Research System

Your multi-agent research system ingests documents from third-party sources — including web scrapes, uploaded PDFs, and external API responses — and passes them to specialized subagents for summarization and fact extraction. Over several weeks, your logs show occasional anomalous external HTTP requests originating from agent processes, which forensic analysis links to instruction-bearing content embedded in processed documents. Your current architecture passes credentials to each subagent at initialization so they can authenticate downstream API calls. What architectural change most effectively reduces the blast radius of a successful prompt injection attack against a subagent?

A. Add explicit instructions to each subagent's system prompt warning it to ignore any directives embedded in document content and to never make unauthenticated external calls.

B. Move credentials outside the subagent's trust boundary by routing all authenticated API calls through an external proxy that injects credentials — subagents request the proxy endpoint rather than holding secrets directly.

C. Add a post-processing validation agent that reviews each subagent's outputs and API call logs before results are propagated to the orchestrator, flagging suspicious patterns for human review.

D. Consolidate the specialized subagents into a single agent with broader context so that no credential handoff is needed between components, eliminating the injection surface at handoff points.