A Claude application is occasionally refusing to answer questions that should be in scope, including questions the application has answered correctly in the past. You want to investigate.
What is the first step of your investigation?

• A. Examine traces of the refused requests to identify what triggers the refusal, whether input patterns, system prompt content, or other context.
• B. Assess whether the refusals cluster around specific input patterns, question types, or time periods by analyzing the distribution of refused requests in the logs.
• C. Test whether removing or adjusting those instructions resolves the behavior.
• D. Identify which questions have been refused and compare them against previously answered questions to determine what changed between the successful and refused interactions.

A teammate has asked you to explain the difference between context engineering and prompt engineering.
They have heard the terms used interchangeably and are unsure how each applies to a Claude application that processes long-running multi-step tasks.
How would you describe the distinction?

• A. Prompt engineering focuses on the model's response, while context engineering focuses on the user's input across many sessions in a long-running multi-step Claude application.
• B. Prompt engineering and context engineering each address content the team gives Claude, but the team can group them under a single workflow because the practices use overlapping techniques.
• C. Prompt engineering is the older term for prompt design, while context engineering is the newer term that has replaced it in modern Claude applications across the industry.
• D. Prompt engineering shapes individual prompts for specific outputs, while context engineering manages how content flows across turns and steps and takes steps to keep relevant state visible.

You are designing a Claude application that will process customer support tickets in two stages: a triage stage that classifies tickets and a response stage that drafts replies. The team is debating whether to use a single Claude call that handles both stages or separate Claude calls for each stage.
How would you structure the application?

• A. Use a single Claude call for both stages, on the grounds that a single call is cheaper than multiple calls in any production Claude application setup.
• B. Use multiple Claude calls in parallel that each draft a complete ticket reply, then have a fourth Claude call select the best one to send to the customer.
• C. Use separate Claude calls for triage and response, because each stage has distinct inputs, outputs, and success criteria that benefit from focused prompts.
• D. Use a single Claude call for triage and then use a non-Claude rule-based system for response generation, on the grounds that rule-based systems are more reliable for drafting replies.

Your Claude application uses tool calling to fetch patient data and generate summary reports. The flow occasionally fails because the model returns a tool_use block that references arguments not present in the schema, and your application code does not handle this case gracefully.
How would you address this?

• A. Retry the same request repeatedly until the model returns a valid tool_use block that matches the schema as expected.
• B. Stop using tool calling entirely and replace tools with prompted text generation that asks the model to describe what it would do.
• C. Log invalid tool_use blocks when they occur and allow the tool dispatch to proceed, relying on the tool's own error handling to surface failures back to the application.
• D. Validate the tool_use block's arguments against the tool schema before dispatching the tool and handle invalid arguments as a recognized error path.

A teammate has asked you to explain why your Claude agent's tools include detailed descriptions in the tool definition, even when the tool name is already descriptive. The teammate suggests removing the descriptions to simplify the tool definitions.
How would you respond?

• A. Suggest moving the descriptions out of the tool definition and into a separate documentation file the team maintains so the tool definitions stay short and the descriptions remain available.
• B. Agree with the teammate because tool names are sufficient for the model to choose the right tool on every request the agent handles.
• C. Explain that the model uses the tool description to decide when to call the tool, and descriptions disambiguate cases where the tool name is not enough.
• D. Suggest replacing the descriptions with example calls embedded in the tool definition, treating example calls as a complete substitute for the prose description.