CCAR-P Practice Test 12 — 15 Questions

Question 1

A financial services firm runs a Claude-powered loan pre-screening agent that reads applicant emails and drafts a preliminary recommendation. The firm's risk committee insists that no automated decision may be finalized without a documented human review step. Which control most directly satisfies this requirement while preserving the agent's throughput?

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

Accepted Answer

Insert a mandatory human-in-the-loop approval gate before any recommendation is written back to the loan origination system.. The risk committee's requirement is about who authorizes a decision, not how the model behaves or how well activity is recorded. Only a human-in-the-loop approval gate makes a person the final authorizing step before the recommendation reaches the loan origination system, yielding the documented review the committee expects while letting the agent handle upstream work automatically.

Answer

Log every model request and response to immutable storage and retain the logs for seven years.

Answer

Increase the model's temperature to zero and add a system prompt instructing Claude to be conservative in its recommendations.

Answer

Enable prompt caching on the applicant-email summarization step to reduce latency and cost per screening.

Question 2

A Claude-based customer onboarding assistant has been in production for one quarter. The steering committee asks how you will decide whether to expand it to two additional regions. Which approach best supports that lifecycle decision?

Accepted Answer

Run a defined evaluation against agreed success metrics from the pilot, then present results as a stage-gate decision to the committee.. Lifecycle decisions at scale should be evidence-based and governed through stage gates. By measuring the pilot against pre-agreed success metrics and presenting the results to the steering committee, you convert an expansion question into a structured go/no-go decision. This protects the organization from scaling unproven value, creates a documented rationale, and ensures regional differences in language, compliance, and support are addressed before commitment.

Answer

Expand immediately to both regions because the assistant has had no reported outages in production.

Answer

Wait until a region formally requests the assistant, then build it on demand.

Answer

Survey the engineering team about whether they feel the assistant is ready for more traffic.

Question 3

A developer productivity team runs an internal Claude-powered code review assistant. During peak morning hours, many developers submit large diffs simultaneously and the assistant returns HTTP 429 responses with a retry-after header. The team wants to keep latency predictable for interactive users while still processing queued batch jobs overnight. Which design best achieves this?

Accepted Answer

Separate interactive and batch traffic into distinct queues with different priority, apply exponential backoff with jitter on 429s, and use retry-after to pace the batch queue.. Traffic isolation plus backoff with jitter and retry-after pacing is the standard pattern for protecting interactive latency under throttling. Batch work can tolerate delay and should absorb the throttling signal, while interactive requests deserve priority. This keeps the developer experience stable without ignoring the server's pacing guidance.

Answer

Immediately retry every 429 request in a tight loop until it succeeds, because the retry-after header is only advisory and can be ignored for interactive traffic.

Answer

Switch all traffic to a smaller, cheaper model during peak hours to reduce the chance of hitting rate limits, then switch back after the peak window ends.

Answer

Increase the max_tokens parameter on every request so fewer round trips are needed, reducing the total number of API calls during peak hours.

Question 4

You are designing a long-running agent that executes a sequence of irreversible operations, such as issuing refunds and sending customer notifications. The agent runs unattended overnight. Which TWO architectural patterns best ensure that a partial failure does not leave the system in an inconsistent state? (Choose two.)

Accepted Answer

Implement compensating actions so that each irreversible operation has a defined reversal or mitigation step.. Unattended agents performing irreversible actions need both a durable record of intent and a way to reverse or mitigate completed steps. Journaling provides the audit trail and restart logic, while compensating actions provide the semantic undo. Together they allow the orchestrator to detect partial completion and bring the workflow back to a consistent state after a crash or timeout.

Answer

Cache all tool responses in memory so the agent can replay the session without re-calling tools after a restart.

Answer

Raise the max_tokens limit for each planning step so the agent can reason through more contingencies before acting.

Answer

Increase the model's temperature so it explores alternative execution orders and avoids getting stuck on a failing step.

Question 5

A software vendor is preparing for a customer security review of its Claude-powered support assistant. The customer asks how the vendor prevents the assistant from leaking one tenant's data into another tenant's conversation. Which architectural control most directly addresses this concern?

Accepted Answer

Scope every retrieval query and every stored conversation to the authenticated tenant identifier, and reject requests whose tenant scope cannot be resolved.. Cross-tenant leakage is prevented at the point where data is selected, not at the point where text is generated. Binding every retrieval and storage operation to the authenticated tenant identifier, and failing closed when that identifier is missing, guarantees that only authorized content can reach the model. Prompt instructions, extended thinking, and temperature settings alter model behavior but leave the underlying data access path unchanged.

Answer

Add a system prompt instructing the assistant to answer only questions about the current customer.

Answer

Set the assistant's temperature to zero to make responses more predictable across tenants.

Answer

Enable Claude's extended thinking so the assistant reasons more carefully before answering.

Question 6

An enterprise architect is designing a Claude deployment where a single prompt may contain data belonging to customers in the EU, Brazil, and California. The legal team requires that each data subject's rights be honored independently and that processing purposes be documented per jurisdiction. Which architectural approach best supports this requirement?

Accepted Answer

Implement data classification and purpose tagging at ingestion so each record carries jurisdiction, lawful basis, and permitted processing purpose metadata that the orchestration layer enforces at prompt assembly time.. Honoring rights independently while documenting purpose per jurisdiction requires machine-readable metadata attached to each record and enforced during prompt assembly. Classification tags for jurisdiction, lawful basis, and permitted purpose let the orchestration layer exclude data whose purpose does not match the current request, which is the only approach here that scales and remains auditable. Centralization, endpoint selection, and manual pruning do not deliver per-subject enforcement or purpose documentation.

Answer

Route all prompts through a single global data lake so that lineage is consistent and easy to report.

Answer

Rely on the model provider's regional endpoints to satisfy all three jurisdictions automatically.

Answer

Ask business users to manually remove data from prompts when it is not needed for the current task.

Question 7

A government agency wants assurance that its Claude deployment will not be used to generate content that violates Anthropic's usage policies. Which action most directly supports ongoing policy compliance?

Accepted Answer

Review Anthropic's usage policies and configure the deployment so its use cases fall within permitted categories, documenting that mapping for internal review.. Policy compliance is achieved by understanding the provider's acceptable use categories, deliberately designing the deployment to stay within them, and documenting that alignment so it can be reviewed. Configuration changes, spending tier, and identity obfuscation do not alter the obligation to use the service within policy and provide no evidence of alignment.

Answer

Disable all safety filters so the agency retains maximum control over what the model is allowed to output.

Answer

Route all requests through a third-party gateway that anonymizes the agency's identity from Anthropic.

Answer

Purchase the highest available usage tier so the agency's traffic is treated as a trusted enterprise workload.

Question 8

A media company uses Claude through the Anthropic API to draft articles. Legal counsel asks the platform team to demonstrate that every published draft can be traced to the exact model behavior that produced it, even after Anthropic deprecates older models. The team currently calls the alias claude-sonnet-4-5. Which change best satisfies counsel's requirement?

Accepted Answer

Pin requests to a dated model snapshot identifier and archive the full request parameters and response with each draft.. Traceability to a specific model behavior requires two things: an immutable model identifier and a durable record of the exact input and output. A dated snapshot pins behavior so it can be reproduced or referenced after deprecation, while archiving request parameters and responses preserves the actual interaction. Aliases, caching, and sampling settings modify cost, latency, or variance but none of them establish which model version produced a given published artifact.

Answer

Increase max_tokens and set temperature to zero so outputs become deterministic.

Answer

Log the alias string claude-sonnet-4-5 alongside each draft in the content management system.

Answer

Enable prompt caching so identical requests return consistent outputs across model updates.

Question 9

A team is rolling out an internal Claude-powered assistant for their engineering organization. Adoption is low and developers report that they do not trust the answers for anything beyond trivial questions. The enablement lead wants to increase adoption by making the assistant's behavior more transparent and debuggable. Which change best supports that goal?

Accepted Answer

Expose the system prompt, the tools available, and the retrieved context used for each answer, and log request IDs for support.. Trust grows when developers can see the inputs that shaped an answer. Exposing the system prompt, available tools, and retrieved context lets engineers verify whether the assistant had the right information and instructions. Logging request IDs enables targeted investigation of bad answers, turning vague distrust into specific, fixable issues that the enablement team can address.

Answer

Restrict the assistant to answering only questions about internal documentation and disable all other capabilities.

Answer

Hide the system prompt and tool definitions from users to keep the interface simple.

Answer

Increase the model's temperature so answers vary more and feel more natural to developers.

Question 10

A fintech startup wants to use Claude to generate marketing copy that references competitor products by name and makes performance comparisons. Legal counsel asks the architect which governance step is most appropriate before this capability goes live.

Accepted Answer

Establish a documented review process that classifies the use case against Anthropic's Usage Policies and applicable advertising regulations before deployment.. The scenario involves a potentially sensitive use case: comparative advertising that names competitors. Governance best practice is to classify the use case against Anthropic's usage policies and relevant advertising regulations through a documented review before deployment. Sampling parameters, proofreading, and caching affect style, quality, and cost, and none of them determine whether the activity is permitted or lawful.

Answer

Increase the model's temperature setting so the copy sounds more creative and varied.

Answer

Enable prompt caching to lower the cost of generating large volumes of marketing variants.

Answer

Add a spell-check and grammar pass to the generated marketing copy before publication.

Question 11

A retail company's legal team discovers that several engineering squads have been calling the Anthropic API with personal API keys obtained on individual credit cards, outside the corporate agreement. Leadership wants a governance model that both eliminates this practice and preserves the ability to audit all Claude usage centrally. Which governance model best achieves this?

Accepted Answer

Provide a centralized Anthropic Console organization with workspace-scoped API keys issued per squad, and route all usage through a corporate gateway that logs every request.. Eliminating shadow AI requires giving teams a sanctioned, easy path while removing the unmanaged one, which a centralized Console organization with workspace-scoped keys accomplishes. Central auditability then follows from routing all traffic through a corporate gateway that records every request. Policies, attestations, blanket blocks, and vendor-side monitoring either lack enforcement teeth, suppress legitimate use, or rely on attribution the vendor cannot make.

Answer

Publish a policy prohibiting personal API keys and require engineers to attest annually that they comply.

Answer

Block outbound traffic to api.anthropic.com at the corporate firewall for all users except the platform team.

Answer

Ask Anthropic account management to monitor for personal keys belonging to company employees and report them monthly.

Question 12

Six months after launch, a logistics company's operations VP reports that the Claude-based exception-handling assistant 'used to be great and now gives worse answers,' though no code has changed. You confirm the application code and system prompt are untouched. What is the most likely explanation you should investigate first?

Accepted Answer

The assistant is now encountering a broader distribution of exception types than during the pilot, including cases that were out of scope for the original evaluation set.. When code, prompt, and model snapshot are unchanged but users report degradation, input distribution drift is the leading hypothesis. Early pilots exercise a narrow set of exception types chosen by the implementation team, while production usage expands into long-tail cases the evaluation never covered. Measuring the current mix of incoming exception categories against the original test set either confirms drift or rules it out, and it is a cheap first check before investigating infrastructure.

Answer

The prompt caching layer has expired its entries, forcing the model to reason without the operational context it previously had.

Answer

Latency has increased over time, and the operations team is now perceiving slower responses as lower quality answers.

Answer

The model provider has silently retrained the underlying model, so previously correct behaviours have been overwritten without notice.

Question 13

A platform team maintains a shared prompt library used by 40 internal services. After a subtle wording change to a summarization prompt caused a 12% drop in a downstream classification F1 score, the team wants every prompt change to be reviewable, version-pinned, and automatically regression-tested before rollout. Which approach best satisfies these requirements?

Accepted Answer

Store prompts as versioned files in a Git repository, require pull-request review, and run a CI job that evaluates each changed prompt against a held-out golden dataset before merge.. Treating prompts as versioned code in Git couples three needed controls: peer review through pull requests, immutable references through commits or tags, and automated quality gates through a CI evaluation against a golden dataset. Because the evaluation runs before merge, a wording change that harms the downstream classification score is blocked rather than discovered in production, which protects all consuming services.

Answer

Move all prompts into a shared spreadsheet with a change log column, and instruct service owners to copy the latest text into their code before each release.

Answer

Deploy every prompt change straight to production behind a feature flag, then compare aggregate F1 scores over the following week and revert manually if quality declines.

Answer

Keep prompts inline in each service's source code and rely on each team's existing unit tests, which mock the Claude response, to catch quality regressions.

Question 14

A customer-support agent must sometimes escalate to a human and sometimes resolve autonomously. The compliance team requires that any action touching billing be reviewed by a human before execution, while password resets may proceed automatically. The architect wants the model to decide routing without hardcoding every rule in the prompt. Which design best satisfies the requirement?

Accepted Answer

Classify each proposed tool call by risk tier in the orchestration layer, auto-approve low-risk actions, and require human approval for billing-tier actions regardless of model output.. The compliance requirement is a hard gate, so the decision to require human approval must live in deterministic orchestration code rather than in model judgment. Tiering tool calls by risk lets low-risk actions like password resets flow automatically while billing actions are held for approval before execution. The model can still propose actions, but the authoritative approval decision is enforced outside it, which makes the control auditable and immune to prompt drift or probabilistic misrouting.

Answer

Fine-tune the model on historical escalations so it learns which actions compliance typically requires humans to approve.

Answer

Give the model a single escalate tool and instruct it in the system prompt to use judgment about when human review is required.

Answer

Log every tool call the agent makes and have the compliance team review the logs weekly to catch any unauthorized billing actions.

Question 15

A platform team is building a shared Claude integration library used by 40 internal microservices. Each service currently hard-codes its own model ID, max_tokens, and retry logic. The team wants a single place to roll out model upgrades and enforce consistent retry behaviour without redeploying every service. What is the most effective architecture for this requirement?

Accepted Answer

Publish a versioned internal SDK that reads model configuration from a central configuration service at startup and exposes typed client wrappers with built-in retry and backoff.. A versioned SDK reading from a central configuration service gives the platform team one lever for model IDs, token limits, and retry policy while preserving per-service autonomy. It enforces consistent behaviour through typed wrappers, supports staged rollouts, and avoids the fragility of repository-level environment files or an opaque rewriting proxy. This is the standard pattern for operational enablement at scale.

Answer

Place an HTTP proxy in front of the Anthropic API that rewrites the model field on every request and centralizes retries.

Answer

Move all Claude calls into a single shared monolith service and have the 40 microservices call it over gRPC.

Answer

Add a shared environment variable file to each repository and require every service owner to pull the latest values before each deployment.