CCAO-F Practice Questions: Troubleshooting and Optimization Domain

A) Ask employees to reformat the output manually before every system entry.

B) Switch to a stronger model because formatting is tied to model capability.

C) Specify the downstream-required structure and validate the output against it.

D) Generate several formats and let users choose whichever one pastes correctly.

A) Increase the number of categories so every comment has more possible destinations.

B) Ask Claude to choose whichever label appears most common in prior outputs.

C) Clarify category definitions and add representative borderline examples.

D) Use a longer explanation for every classification while keeping definitions unchanged.

A) Continue adding corrective messages until the old criteria stop appearing.

B) Ask Claude to remember which assumptions are obsolete without restating them.

C) Start a clean context using a verified summary of the current requirements.

D) Switch models and retain the full conversation unchanged, even if that creates some additional manual effort.

A) Keep each step manual so users retain maximum control over routine work.

B) Standardize the repeated sequence into a reusable workflow while preserving needed review points.

C) Ask users to skip whichever step seems least important each time.

D) Use a stronger model so the number of workflow steps no longer matters.

A) Add more source material so Claude better understands the customers.

B) Increase the model capability so the replies become more comprehensive.

C) Keep the workflow because draft speed is already good, even if that creates some additional manual effort.

D) Revise the output instructions to target only the information reviewers actually need.

A) Remove review first, then restore it only if a serious error occurs.

B) Evaluate which reviews are required by risk, policy, and observed failure modes before changing them.

C) Replace all human review with Claude self-checks because they are faster, even if that creates some additional manual effort.

D) Remove review from the highest-risk outputs first because those create the most delay.

A) Ask reviewers to continue correcting the mistake because they already know the difference.

B) Increase the number of comparison fields so the distinction becomes more obvious.

C) Clarify the definitions and provide an example showing how the two fields differ.

D) Use research mode so external information can resolve the terminology.

A) Continue copying them because repeated manual setup prevents stale configuration.

B) Ask each employee to create their own shorter version of the instructions.

C) Use a stronger model so fewer instructions are needed in each chat.

D) Move the stable recurring guidance into an appropriate persistent configuration.

A) Keep the sections because more complete reports are always preferable.

B) Ask the reviewer to create a macro that deletes the sections faster.

C) Use a different model that tends to write shorter reports.

D) Remove the unused sections from the required output specification.

A) Switch to a stronger model because structure inconsistency indicates low capability.

B) Provide an explicit schema or template with required field names and formats.

C) Ask Claude to explain each exception in prose before producing the records.

D) Generate three versions and keep whichever structure appears most complete.