CCAR-P Practice Test 14 — 15 Questions

Question 1

A startup is using Claude to generate marketing copy. The legal team is concerned about potential copyright infringement if the model reproduces copyrighted text. Which governance measure should the startup implement to best mitigate this risk?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Use a plagiarism detection tool to scan all generated content before publication.. The plagiarism detection tool is the most direct mitigation because it actively checks generated content against known copyrighted works and flags potential matches. This allows human review and modification before publication. Other options either do not prevent infringement or are ineffective, such as disclaimers or temperature adjustments.

Answer

Include a disclaimer that any resemblance to existing works is coincidental.

Answer

Set the model's temperature to zero to ensure deterministic outputs.

Answer

Fine-tune Claude on the startup's own proprietary content to avoid using external data.

Question 2

A production agent uses the Claude Messages API with extended thinking enabled to solve multi-constraint scheduling problems. The agent must preserve its reasoning across several tool calls within a single user turn. A developer notices that after the second tool call, the model appears to forget earlier constraints it had already reasoned about. Which change best preserves the reasoning chain across tool calls?

Accepted Answer

Pass the full assistant message, including thinking blocks, back into the conversation history on each subsequent request.. Extended thinking stores reasoning in thinking blocks that are part of the assistant message. When an agent makes multiple tool calls within one turn, each follow-up request must include the prior assistant message with its thinking blocks intact. Omitting them causes the model to lose the constraints it already reasoned through, producing the observed regression after the second tool call.

Answer

Disable extended thinking for subsequent requests once the first tool call has returned, since the constraints are already known.

Answer

Increase the thinking budget_tokens value so the model has more room to reason on each request.

Answer

Summarize the thinking blocks into a single text note and append it as a user message before the next tool call.

Question 3

A team maintains a library of internal Claude prompts used by several services. They want to update a shared system prompt once and have every service pick up the change without redeploying, while keeping a rollback path if quality regresses. Which approach best satisfies both requirements?

Accepted Answer

Serve prompts from a versioned prompt registry that services fetch at runtime, with the ability to pin or roll back to a prior version.. The requirements point to treating prompts as versioned runtime data rather than code. A prompt registry lets the team publish once, have services fetch the current version, and revert by repointing to a prior immutable version. Duplicated prompts, shared-library constants, and host-level environment edits each force redeploys, lack reliable rollback, or both, so they cannot meet the stated goals.

Answer

Copy the system prompt into each service's repository and coordinate releases manually.

Answer

Keep the system prompt in an environment variable that operators edit directly on running hosts.

Answer

Embed the system prompt as a constant in a shared library package and publish a new package version for each change.

Question 4

You are building a Claude-based agent that must parse unstructured customer emails, extract line-item order data, and then call a fulfillment tool with the extracted values. During testing, the agent occasionally calls the fulfillment tool with empty or garbled line items when an email contains a forwarded message with a different formatting style. Which architectural change most directly reduces this failure?

Accepted Answer

Add a validation step that checks the extracted line items against the tool's JSON schema and returns a structured error to the model before any tool call is allowed.. The failure occurs because extracted line items are not verified before the tool boundary is crossed. A validation gate that compares the extraction against the tool's JSON schema and returns a structured error gives the model a chance to repair its output. This keeps invalid data out of downstream systems and converts a silent corruption into a retryable, observable event.

Answer

Reduce the number of tools available to the agent so it focuses only on fulfillment during extraction.

Answer

Switch the fulfillment tool to accept free-form natural language instead of structured parameters so the model can pass whatever it extracted.

Answer

Increase the max_tokens parameter on the extraction call so the model has more room to emit complete line items.

Question 5

A media company uses Claude to moderate user-generated comments at high volume. The risk team wants a control that detects when the moderation model's behavior drifts, for example becoming unusually permissive or aggressive, before it affects the community at scale. Which control best fits this need?

Accepted Answer

Maintain a labeled evaluation set of representative comments and run scheduled regression tests that compare current moderation decisions against expected outcomes, alerting on threshold breaches.. Detecting behavioral drift requires a stable reference: a labeled evaluation set that defines expected moderation outcomes. Running scheduled regression tests against it converts drift into a measurable deviation and supports alerting before community-wide impact. Token monitoring, user appeals, and verdict caching each serve different purposes and cannot reveal systematic changes in how the model judges content.

Answer

Provide an appeal button so users can report comments they believe were moderated incorrectly.

Answer

Monitor average tokens per moderation request to detect changes in comment length.

Answer

Cache moderation verdicts for identical comments to reduce repeated inference costs.

Question 6

A team wants an agent to answer questions about a large internal corpus. They notice the agent invents details when the retrieved chunks are only loosely related to the question. Which change most directly reduces fabricated answers grounded in weak evidence?

Accepted Answer

Instruct the agent in the system prompt to answer only from provided documents and to state when evidence is insufficient.. The reported failure is generation overreach under weak retrieval, so the highest-leverage fix is a grounding instruction that scopes answers to supplied documents and explicitly permits declining. This directly constrains what the model may assert and gives it a safe alternative to invention, independent of any retrieval tuning that might follow.

Answer

Raise the temperature setting so the model explores more diverse phrasings of the answer.

Answer

Increase the number of retrieved chunks returned by the vector search to fifty per query.

Answer

Switch the embedding model to one with a larger vector dimension for finer similarity scoring.

Question 7

Midway through a Claude deployment, the executive sponsor asks to add a new capability that would require sending customer records to a third-party retrieval service outside the approved environment. Which action best demonstrates sound stakeholder communication and lifecycle governance?

Accepted Answer

Acknowledge the request, document the proposed data flow, assess privacy and compliance impact with the relevant reviewers, and return a decision with options and tradeoffs.. The sound response acknowledges the sponsor's request, makes the proposed data flow explicit, involves privacy and compliance reviewers, and returns a decision with options and tradeoffs. This preserves both the relationship and the control environment, and it gives the sponsor a genuine choice. Immediate approval, outright rejection, and building first each bypass a review step that the environment requires.

Answer

Reject the request outright because the original scope did not include third-party retrieval services.

Answer

Approve the addition immediately to preserve sponsor satisfaction, and document the data flow change after implementation.

Answer

Implement the capability in a personal development environment to demonstrate feasibility before raising the governance question.

Question 8

Your organization is standardizing a Claude-based internal assistant across three departments with different risk tolerances and workflows. Executive sponsors want a single governance model that keeps the program aligned and auditable as it grows. Which TWO practices should you establish as part of the lifecycle governance? (Choose two.)

Accepted Answer

Maintain a central registry of approved use cases, model versions, and evaluation results that all departments reference and update as the program evolves.. Effective lifecycle governance balances consistency with flexibility. A risk-tiered change-control process ensures high-impact modifications receive review while routine iteration stays fast, and a central registry provides shared visibility into approved use cases, models, and evaluation evidence. Together they let three departments with different needs operate under one auditable model without freezing progress or forcing one-size-fits-all artifacts.

Answer

Freeze all prompts and model versions for twelve months so that behavior cannot change while the governance model is being validated.

Answer

Let each department maintain its own independent prompt library, evaluation approach, and release cadence with no shared standards or reporting.

Answer

Require every department to use identical prompts and thresholds regardless of their differing workflows and risk tolerances.

Question 9

You are the lead architect for a Claude-based claims triage assistant at an insurance company. Two weeks before the pilot goes live, the Head of Compliance asks how they will be able to demonstrate, months later, which version of the system prompt and which model snapshot produced a given claim recommendation. What should you implement to meet this requirement?

Accepted Answer

Emit a structured audit record per recommendation containing the model snapshot ID, system prompt version hash, parameters, and a request identifier, and store it immutably.. Compliance needs retrospective, per-decision traceability, which requires capturing execution metadata at inference time: the model snapshot identifier, a hash of the system prompt version, the parameters used, and a correlation identifier tied to the claim. Immutable storage keeps that evidence tamper-evident and queryable months later. Reasoning traces and repository tags describe intent or internal deliberation but never bind a specific recommendation to the exact deployed configuration.

Answer

Version the system prompt in Git and tag each release, then tell Compliance that the tag history is the authoritative record of what was deployed.

Answer

Rely on the Anthropic Console usage logs, which retain every request and response for the account and can be queried retroactively by claim ID.

Answer

Enable extended thinking on the triage calls and archive the returned thinking blocks alongside each claim recommendation.

Question 10

Six months into production, a Claude-based claims-triage system shows a slow decline in acceptance of its recommendations, from 82 percent to 61 percent, with no code or model changes. The operations manager asks what to do. Which investigation best addresses the root cause?

Accepted Answer

Segment acceptance by claim type, region, and time, and review samples of recently rejected recommendations with the adjusters who rejected them.. When a stable system degrades without code or model changes, the likely drivers are input drift, population change, or evolving human judgement. Segmenting the metric localizes the problem, and reviewing rejected recommendations with the adjusters reveals the mechanism behind each rejection. Retraining, rolling back an unchanged version, or passively monitoring all bypass that diagnosis and either risk worsening the issue or prolonging it.

Answer

Roll back to the model version that was in production when acceptance was 82 percent.

Answer

Immediately retrain or fine-tune the model on the most recent accepted claims to restore the previous acceptance rate.

Answer

Report the metric to the steering committee as a normal variance and continue monitoring for another quarter.

Question 11

A new engineer joins a team building a Claude-powered support triage tool. She wants to iterate quickly on prompt wording without redeploying the service, but the team also needs every prompt change to be auditable and reversible. Which practice best supports both goals?

Accepted Answer

Store prompts in an external configuration store with version history, and have the application load the active prompt version at runtime with a rollback control.. Externalizing prompts into a versioned configuration store decouples prompt iteration from code deployment while preserving an auditable history and a fast rollback path. The engineer can experiment and promote versions without a build pipeline, and the team can always trace or revert a change, which satisfies both stated goals.

Answer

Keep prompts in a shared document that the engineer edits freely, and copy the current text into the service configuration manually before each release.

Answer

Hard-code the prompt as a string literal in the application source and require a full deployment for every wording change.

Answer

Allow the engineer to edit prompts directly in the production database with a direct SQL client, keeping no history of prior versions.

Question 12

A developer is building an internal tool that uses Claude to answer questions about a large codebase. They want to reduce hallucinations and ensure answers are grounded in the actual code. Which technique should they use?

Accepted Answer

Use retrieval-augmented generation (RAG) by embedding the codebase and injecting relevant snippets into the prompt.. Retrieval-augmented generation retrieves relevant code snippets and includes them in the prompt, grounding the model's answers in actual code. This reduces hallucinations and keeps responses up-to-date as the codebase evolves. It is more scalable and cost-effective than fine-tuning or stuffing the entire codebase into the context window.

Answer

Use a larger context window to include the entire codebase in every prompt.

Answer

Fine-tune the model on the entire codebase.

Answer

Increase the model's temperature to encourage more creative answers.

Question 13

A multinational bank runs a Claude-powered assistant that drafts internal credit memos. Auditors require the bank to prove that every generated memo can be traced to the exact human who requested it, the data sources the model retrieved, and the decision it influenced. Which governance capability should the architect implement first to satisfy this requirement?

Accepted Answer

Enable structured audit logging that records request identity, retrieved source references, and downstream model output identifiers for each Claude invocation.. The requirement is evidentiary lineage: for each generated memo, the bank must show the requesting human, the retrieved sources, and the output that influenced a decision. Structured audit logging is the only control that binds identity, retrieval context, and output reference into a reviewable record. Filtering, caching, and spend dashboards address confidentiality, efficiency, and cost respectively, none of which reconstruct the chain of custody auditors demand.

Answer

Configure automatic prompt caching to reduce latency and cost for repeated credit memo templates.

Answer

Deploy a content moderation layer that filters sensitive financial terms before prompts are sent to Claude.

Answer

Set up a dashboard that tracks aggregate token consumption and monthly API spend per business unit.

Question 14

A multinational insurer wants to quantify how often its Claude-based claims assistant produces outputs that violate its internal tone and fairness policy before it expands the pilot to three new countries. The compliance team needs a repeatable, statistically defensible measurement rather than anecdotal review. Which approach best meets this need?

Accepted Answer

Run a structured evaluation over a stratified sample of representative claims, scoring each output against a written tone-and-fairness rubric with two independent raters.. Quantifying violation frequency requires a defined sample, a consistent scoring instrument, and controls for scorer bias. Stratified sampling over representative claims plus a written rubric scored by two independent raters yields a rate with reliability evidence, which is what lets compliance compare countries and justify or reject expansion on evidence rather than impressions.

Answer

Review the model's system prompt and knowledge sources with legal counsel to confirm the policy language is correctly worded.

Answer

Monitor the assistant's average response latency and error rate in production dashboards during the pilot period.

Answer

Ask the pilot's ten claims adjusters to report any tone or fairness problems they notice in daily use.

Question 15

You are preparing a kickoff briefing for a business unit adopting a Claude-based internal knowledge assistant. The sponsor asks what they should expect in the first 30 days. Which framing best sets accurate expectations and supports lifecycle alignment?

Accepted Answer

The first 30 days will focus on use-case validation, data readiness, evaluation baselines, and defining success metrics with the sponsor.. A credible first-30-days plan centers on validating the use case, assessing data readiness, establishing evaluation baselines, and agreeing on success metrics with the sponsor. This is honest about what determines outcomes and creates shared accountability. Claiming instant production quality, deferring all work to procurement, or promising a fine-tuned model each sets expectations the engagement cannot reliably meet.

Answer

The first 30 days will be spent negotiating the final contract and procurement terms before any technical work begins.

Answer

The system will reach full production quality immediately because Claude is a pre-trained foundation model.

Answer

The first 30 days will deliver a fully tuned model fine-tuned on the business unit's proprietary corpus.