CCDV-F Prompt and Context Engineering Practice Question
A developer is building a customer-support assistant that must answer only from a provided knowledge base article and must refuse to answer when the article does not contain the information. The assistant currently invents plausible answers. Which prompt change best enforces the refusal behavior?
⚠ Common exam trap
The trap here is believing that reordering or enlarging the context will suppress hallucination, when only an explicit grounding and refusal instruction changes the model's behavior.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Instruct the model to answer only using information found in the article and to respond with a fixed phrase such as 'I don't have that information' when the article is insufficient.
Grounding requires two explicit constraints: define the permitted source and define what to do when that source is inadequate. Naming the article as the only allowed source, plus a fixed refusal phrase for missing information, gives the model an unambiguous rule and produces a consistent, detectable abstention. Tone, token limits, and document placement do not establish either constraint.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Instruct the model to answer only using information found in the article and to respond with a fixed phrase such as 'I don't have that information' when the article is insufficient.
Why this is correct
Explicitly scoping the answer to the provided article and defining a concrete refusal phrase gives the model a clear, testable behavior. The fixed phrase makes abstention detectable and consistent, which is exactly what the assistant needs when the knowledge base lacks an answer. It directly constrains the source of truth and the fallback response.
- ✗
Increase the max_tokens parameter so the model has more room to explain its reasoning.
Why it's wrong here
Max tokens controls output length, not factual grounding. Allowing longer responses gives the model more space to elaborate and can even increase the opportunity to fabricate details. Nothing about a larger output budget tells the model to restrict itself to the article or to refuse when the article is insufficient.
- ✗
Move the knowledge base article to the end of the prompt so it is the last thing the model reads.
Why it's wrong here
Positioning the article last can influence attention, but it does not by itself establish an abstention rule. The model may still answer from its own knowledge when the article is silent. Without an explicit instruction to refuse, placement alone cannot enforce the required behavior, so this change is insufficient for the scenario.
- ✗
Add an instruction to answer in a polite and professional tone at all times.
Why it's wrong here
Tone instructions affect style, not grounding. The model can be perfectly polite while still fabricating an answer, because politeness does not constrain where the content comes from. The scenario requires the assistant to abstain when the article lacks the information, and a tone directive provides no mechanism for that decision.
About these practice questions
Courseiva writes every CCDV-F question from scratch — 257 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Anthropic exam blueprint
This CCDV-F practice question is part of Courseiva's free Anthropic certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the CCDV-F exam.