CCAO-F questions and answers — Domain 7: Troubleshooting and Optimisation

6 practice questions on Troubleshooting and Optimisation, the domain weighted at 10% of the CCAO-F exam. Every answer explains why the distractors fail — that is the part that transfers to questions you have not seen before.

Question 1

An output ignores one instruction while following the others. What is the most likely cause?

1. The model is not capable enough
2. The instruction is buried among many others, or it conflicts with another instruction
3. The conversation is too short
4. The wrong feature was used

Answer: B. Instructions compete for attention, and two that contradict each other get resolved arbitrarily. Separating them, making the critical ones prominent, and removing the contradiction is the fix — not a heavier model.

Question 2

Claude declines a legitimate professional request because the phrasing reads as something else. What is the best response?

1. Rephrase the request to disguise its purpose
2. Restate the professional context and purpose plainly
3. Switch to a different model
4. Submit the request repeatedly

Answer: B. Ambiguous phrasing is the usual cause, and stating the actual professional context resolves it. Disguising the purpose is the wrong instinct entirely — if the underlying request is legitimate it does not need concealing, and if it needs concealing it is not legitimate.

Question 3

Which habit most improves efficiency for someone doing similar tasks daily?

1. Using the most capable model for everything to avoid rework
2. Saving prompts that worked, and promoting recurring ones into Project instructions
3. Keeping one long conversation for all work
4. Asking shorter questions

Answer: B. Capturing what worked is what turns a one-off good result into a repeatable one. A heavier model for routine work is slower and consumes limits faster; one long conversation for everything degrades and cannot be reused; shorter questions usually mean less context and worse output.

Question 4

A user's first instinct when output disappoints is to press regenerate. Why is this the weakest available move?

1. It uses too much of the usage allowance
2. The same prompt with the same gaps produces the same class of answer
3. It produces shorter output
4. It resets the conversation history

Answer: B. Regeneration re-rolls the same input, so the same missing context produces the same kind of shortfall. Diagnosing what was absent and supplying it addresses the cause. Allowance consumption is a real but secondary concern.

Question 5

When is switching to a more capable model genuinely the right fix?

1. Whenever output quality disappoints
2. When the task itself is more demanding than the tier you chose can handle, and the prompt is already sound
3. When responses are too short
4. When the conversation has grown long

Answer: B. Model tier is the right lever only once you have ruled out the prompt, and only when the task genuinely exceeds the tier. Short responses are a format instruction; a long conversation is a context problem solved by a fresh chat or a Project.

Question 6

How should a team turn individual troubleshooting lessons into lasting improvement?

1. Rely on people remembering what worked
2. Maintain a shared prompt library and encode recurring corrections into Project instructions
3. Have everyone use identical wording in all prompts
4. Escalate every problem to an administrator

Answer: B. A fix that stays in one person's head is worth almost nothing to the team. A shared library plus Projects that encode the corrections people kept having to make is what converts individual learning into a standard. Identical wording is neither achievable nor the point.

← Previous: Questions & answers — Domain 6  |  Next: Facts to memorise →