Q44 of 60 — CCA Practice Exam

Multi-Agent Research System

Your multi-agent research system uses a classification subagent to route incoming research queries to specialized agents (literature review, data analysis, or expert synthesis). After three weeks in production, you notice that 18% of queries are being rerouted after initial assignment, and the synthesis agent frequently complains that it receives queries better suited for the literature review agent. Examining the classification subagent's prompt, you find it returns only a single <intent> tag with the assigned category — no reasoning. What change to the classification subagent's prompt design would most effectively reduce the rerouting rate?

A. Increase the number of few-shot examples in the classification prompt to cover more query types, especially edge cases where literature review and synthesis overlap, so the model has more patterns to match against.

B. Instruct the classification subagent to output its reasoning inside <reasoning> tags before the <intent> tag, so downstream agents and human reviewers can audit why each query was routed and identify systematic misclassification patterns.

C. Expand the system instructions to tell the subagent to "be more careful and precise when distinguishing between literature review and synthesis queries," reinforcing the importance of accuracy before it commits to a label.

D. Replace the single-label classification scheme with a confidence score for each category, routing each query to whichever agent receives the highest score, so ambiguous queries are handled more gracefully than with a hard label.