Agentic Architecture & Orchestration (D1)
Design agent loops, decompose tasks, control execution.
Agentic Architecture & Orchestration (D1) is about structuring an agent that actually finishes the job: the loop it runs, how it decomposes a goal into steps, and where you keep control of execution. The CCA-Foundations exam tests whether you can reason about agentic loops, coordinator–subagent patterns, and the trade-off between letting Claude drive and constraining it.
Expect questions on designing autonomous task loops, orchestrating multi-agent systems, configuring subagent invocation and context passing, enforcing multi-step workflows with handoffs, applying Agent SDK hooks to intercept tool calls, and managing session state, resumption, and forking — plus the failure modes (runaway loops, lost context, premature termination) and the design choices that prevent them.
This domain spans 7 subtopic areas, covered by 164 practice questions across 36 easy, 85 medium, and 43 hard items.
Your support agent calls process_refund for a $240 charge. The model's response contains an assistant message with refund-confirmation text and a tool_use block invoking escalate_to_human, and stop_reason is 'tool_use'. Your loop must decide what to do before the customer sees anything. What is the correct continuation?
What's covered
Subtopic areas in Agentic Architecture & Orchestration, drawn from the exam blueprint: