Question 3
CCA-F is a newly released exam. These practice questions reflect our best current understanding — the exact wording and question formats may change in the live exam until the certification matures. Use them to learn the concepts, not to memorize.
Scenario: Multi-agent Research System - A document analysis subagent frequently fails when processing PDF files: some have corrupted sections that trigger parsing exceptions, others are password-protected, and sometimes the parsing library hangs on large files. Currently, any exception immediately terminates the subagent and returns an error to the coordinator, which must decide whether to retry, skip, or fail the whole task. This causes excessive coordinator involvement in routine error handling. What architectural improvement is most effective? Which improvement is most effective?
- ACreate a dedicated error-handling agent that monitors all failures via a shared queue and decides recovery actions, sending restart commands directly to subagents.
- BConfigure the subagent to always return partial results with a success status, embedding error details in metadata; the coordinator treats all responses as successful.
- CMake the coordinator validate all documents before sending them to the subagent, rejecting documents that might cause failures.
- DImplement local recovery in the subagent for transient failures and escalate to the coordinator only errors it cannot resolve, including attempted steps and partial results.