CCARF Practice Question

Paper 3 · Question 2

D1 Agentic Architecture & Orchestration · Paper 3

[Scenario 1: Customer Support Resolution Agent] You are building a customer support resolution agent with the Claude Agent SDK. It handles high-ambiguity requests (returns, billing disputes, account issues) via custom MCP tools — get_customer, lookup_order, process_refund, escalate_to_human — targeting 80%+ first-contact resolution while knowing when to escalate. The agent calls lookup_order and your code executes it successfully. How does the model actually receive the order data so it can reason about the next action?

• A.The MCP server pushes the result directly into the model's memory
• B.The model re-reads the tool's description, which now includes the result
• C.Nothing is required — the API executes tools and incorporates results automatically
• D.Your code appends a tool result block (referencing the tool call's ID) to the conversation history and sends the updated messages back to Claude in the next request

Show answer & explanations

Correct answer: D

A: and C describe automatic mechanisms that do not exist — the API does not execute your tools

B: confuses static tool definitions with runtime results

C: This option does not best satisfy the scenario requirements.

D: Tool execution is the harness's job: the loop appends the tool result to the conversation and re-invokes the model. That growing history is the only channel through which the model sees what tools returned.

Related Agentic Architecture & Orchestration questions

More: Paper 3 · All practice questions