A credit-union tester claims “the chatbot is magic so we can skip prompt care, and the API tool already has prompts baked in.” What is the best correction?
Select an answer to reveal the explanation.
Short Explanation
Neither a chat pane nor an API job is a magic vending machine. Fuzzy prompts still yield fuzzy testware in both models. Clear, specific instructions remain required—even when some wording lives inside a tool.
Full Explanation
Both chatbots and LLM-powered testing applications still need clear, specific prompts for relevant testware. A credit-union tester who claims “the chatbot is magic so we can skip prompt care, and the API tool already has prompts baked in” misunderstands that baked-in prompts are still prompts—and still require design, review, and maintenance. Civic finance testware quality tracks prompt quality in either interaction model.
Claims that only chatbots need prompts while API-integrated tools never depend on prompt design fail because applications embed prompt templates or instructions that shape every automated draft. Claims that only API tools need prompts while chatbots produce exam-ready cases from empty messages fail the other direction. Claims that prompt quality is optional in Chapter 1 because neuron coverage replaces instructions invent a CT-AI substitute that does not erase prompting on this syllabus.
Exam caveat: interaction model changes the delivery channel, not the need for clear, specific prompting. “Baked in” ≠ “no prompt.”
Operational check: review chatbot prompt patterns and the API tool’s stored prompt templates for the credit-union suite; require specificity (basis, constraints, format); add prompt review to the municipal Definition of Done for GenAI-assisted testware.