Q2 of 60 — CCA Practice Exam

Multi-Agent Research System

Your multi-agent research system uses a shared memory store to persist intermediate findings across sessions. Research subagents retrieve web content and third-party API responses, summarize them, and write summaries back into the store. After three weeks in production, you discover that several downstream synthesis sessions have been generating reports with fabricated citations and distorted source claims — and the corrupted content traces back to the memory store itself. What is the most likely architectural cause of this failure?

A. The memory store's per-file 100KB cap caused subagents to truncate source summaries, losing attribution data that the synthesis agent then reconstructed incorrectly as fabricated citations.

B. The shared memory store was attached with the default read_write access, so a prompt injection in fetched web or third-party API content could write attacker-controlled material into the store, which later sessions then read as trusted memory.

C. The synthesis agent's session instructions did not explicitly direct it to cross-check memory store contents against original sources before incorporating them into reports, allowing distorted summaries to pass through unchallenged.

D. The memory store was attached to multiple concurrent research subagent sessions simultaneously, causing uncoordinated write conflicts between subagents that silently corrupted the stored summaries over time.