CCARF Practice Question

Paper 3 · Question 23

D1 Agentic Architecture & Orchestration · Paper 3

[Scenario 3: Multi-Agent Research System] A coordinator agent built on the Claude Agent SDK delegates to specialized subagents — web search, document analysis, synthesis, and report generation — to research topics and produce comprehensive, cited reports. The coordinator currently invokes the web search subagent, waits for completion, then invokes the document analysis subagent — doubling research latency even though the two tasks are independent. How do you run them in parallel?

• A.Have the coordinator emit both Task tool calls in a single response, which spawns the subagents in parallel
• B.Ask each subagent to work faster in its system prompt
• C.Run two copies of the entire system and merge the databases
• D.Parallelism is impossible — subagents must always run sequentially

Show answer & explanations

Correct answer: A

A: Parallel spawning is achieved by emitting multiple Task tool calls in one coordinator response rather than across separate turns — independent workstreams then execute concurrently.

B: changes nothing about scheduling

C: is massive infrastructure duplication for a one-line orchestration fix

D: is false — parallel Task calls are a documented pattern

Related Agentic Architecture & Orchestration questions

More: Paper 3 · All practice questions