Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 2

Question 10

A logistics coordination team is advancing Task-level acceptance rubric beyond its proof of concept. Engineers expect that the evaluation process can define observable criteria for a usable output rather than aesthetic fluency. What evidence of setup readiness should be required first?

  1. Arrange agreed pass-fail examples and acceptance threshold
  2. Arrange gold labels including unanswerable requests
  3. Arrange reference documents and a citation validation rubric
  4. Arrange policy test harness and action audit events

Explanation

Correct Answer

A:

The proposed design needs agreed pass-fail examples and acceptance threshold as its foundation. This matters because the evaluation process can define observable criteria for a usable output rather than aesthetic fluency; a plausible demonstration without the prerequisite can still leave a case where generic satisfaction scores may reward confident errors. The relevant check is expert-reviewed acceptance rate by task category.

Incorrect Answers

B:

The setup is useful to enable the alternative's behavior where the evaluation process can evaluate correctness jointly with justified abstentions across uncertainty levels and incomplete evidence should yield escalation rather than invented certainty; but the production dependency for Task-level acceptance rubric is agreed pass-fail examples and acceptance threshold.

C:

Where the system summarizes regulated policies with citations, the evaluation process can score whether factual claims are supported by cited source spans; for this release, the team must first verify agreed pass-fail examples and acceptance threshold.

D:

For the separate challenge where an assistant can invoke sensitive operations through tools, the evaluation process can measure disallowed actions that reach a protected side-effecting endpoint; however, the named intervention needs agreed pass-fail examples and acceptance threshold before its protection can be enforced.

Question 11

For a contract administration rollout, the proposed design uses Workflow observation interviews, relying on the claim that the delivery team can observe actual exception handling and frontline work before mapping automation. What enabling control must be in place to make that design enforceable?

  1. Establish cross-functional workflow mapping sessions
  2. Establish stakeholder register and governance charter
  3. Establish baseline metrics and defined evaluation sample
  4. Establish access to operators and representative case samples

Explanation

Correct Answer

D:

Readiness depends on access to operators and representative case samples, not just a design approval. That condition supports the mechanism in which the delivery team can observe actual exception handling and frontline work before mapping automation. Without it, executive summaries alone omit edge cases that determine project success; monitoring discovered exceptions covered by the agreed requirements helps verify the deployed result.

Incorrect Answers

A:

For the separate challenge where automation is proposed without mapping present handoffs and failure paths, the delivery team can chart the existing input-to-outcome process and exception volumes; but the production dependency for Workflow observation interviews is access to operators and representative case samples.

B:

Where product, security and legal teams disagree on who can approve release, the delivery team can define accountable decision owners and consultation rights early; for this release, the team must first verify access to operators and representative case samples.

C:

Where stakeholders ask for a system that is accurate enough without specifying thresholds, the delivery team can turn goals into testable task-specific, risk-sensitive success criteria; however, the named intervention needs access to operators and representative case samples before its protection can be enforced.

Question 12

The solution design for energy operations assumes that the control layer can classify untrusted inputs and prevent them from acquiring instruction authority. Before treating Ingress content classification as production-ready, which dependency needs an owner and verification?

  1. Arrange reviewer authority, timeout path and approval evidence
  2. Arrange documented trust boundaries and security test cases
  3. Arrange recipient entitlement map and redaction tests
  4. Arrange central policy service and auditable subject identity
  5. Arrange authenticated tenant attribution and usage telemetry

Explanation

Correct Answer

B:

The required preparation is documented trust boundaries and security test cases. It makes the following architecture claim enforceable: the control layer can classify untrusted inputs and prevent them from acquiring instruction authority. If that setup is skipped, prompt-only guidance cannot neutralize malicious external text, so the release should examine prompt-injection success rate against controlled attack fixtures.

Incorrect Answers

A:

For the separate challenge where proposed external transfers could produce irreversible customer harm, the control layer can stop before execution and require authorized human confirmation; for this release, the team must first verify documented trust boundaries and security test cases.

C:

The setup is useful to enable the alternative's behavior where the control layer can enforce recipient-specific redaction and validated output fields and generated summaries may echo sensitive identifiers unnecessary for recipients; but the production dependency for Ingress content classification is documented trust boundaries and security test cases.

D:

For the separate challenge where assistants propose privileged write operations in customer systems, the control layer can require server-side authorization and business-policy checks before each side effect; for this release, the team must first verify documented trust boundaries and security test cases.

E:

The setup is useful to enable the alternative's behavior where the control layer can enforce bounded usage budgets and anomaly controls per tenant and one customer can monopolize shared model capacity through abusive retries; however, the named intervention needs documented trust boundaries and security test cases before its protection can be enforced.

Question 13

A program architect at procurement services has endorsed Small-model first-pass routing. The proposed technical effect is that a fast adequate tier handles easy cases and escalates only uncertain ones. Which implementation condition is necessary to realize that effect?

  1. Prepare long-context benchmark cases and token accounting
  2. Prepare measured task complexity and escalation thresholds
  3. Prepare benchmark cases with multi-hop evidence
  4. Prepare representative examples and equalized test settings

Explanation

Correct Answer

B:

The prerequisite creates the necessary operating boundary: measured task complexity and escalation thresholds. The system is intended to work because a fast adequate tier handles easy cases and escalates only uncertain ones. A release without that foundation may still suffer always using the strongest model increases cost and wait time; use routed quality at the target cost and p95 latency to evaluate the result.

Incorrect Answers

A:

This prerequisite supports a method used when legal analysis requests require unusually long source packets, because the model configuration can test effective context utilization and recall before choosing a model configuration; however, the named intervention needs measured task complexity and escalation thresholds before its protection can be enforced.

C:

The setup is useful to enable the alternative's behavior where a stronger model can justify tradeoffs across heterogeneous evidence and multi-document arbitration requires several interdependent inferences; for this release, the team must first verify measured task complexity and escalation thresholds.

D:

Where stakeholders choose a model based on brand rather than target workload, the model configuration can evaluate candidate models on identical task data and operating constraints; however, the named intervention needs measured task complexity and escalation thresholds before its protection can be enforced.

Question 14

A compliance assurance incident exposes a specific weakness: broad tool visibility increases selection error and attack surface. Which architectural correction addresses the cause rather than masking the symptom?

  1. Prioritize Progressive tool discovery
  2. Prioritize Task-scoped tool allowlist
  3. Prioritize Tool ownership and lifecycle review
  4. Prioritize Read-write capability separation

Explanation

Correct Answer

B:

The reported failure calls for the control under which the integration can expose only tools needed for the active task and user permission scope. A prompt-only reminder or unrelated metric would not address the risk that broad tool visibility increases selection error and attack surface. Evidence of resolution requires approved enabling configuration and accountable owner and an improvement in unnecessary tool-call rate and task success under the reduced set.

Incorrect Answers

A:

This addresses the failure where eager inclusion of the entire catalog wastes context and confuses selection, because the integration can advertise compact discovery metadata and load detailed definitions on demand; but this remedy operates on another boundary, not the case where broad tool visibility increases selection error and attack surface.

C:

This addresses the failure where overlapping names lead to arbitrary selection and inconsistent answers, because the integration can retire or consolidate redundant capabilities with a responsible owner; yet it would not correct the observed weakness: broad tool visibility increases selection error and attack surface.

D:

When mixing reading and mutation in one tool overgrants routine sessions, the integration can expose read-only access independently from privileged changes; but this remedy operates on another boundary, not the case where broad tool visibility increases selection error and attack surface.

Question 15

The software platform deployment uses Outcome-first use-case framing, but executive reports currently count only successful model responses. The residual concern is that feature demonstrations can obscure the absence of an accountable success condition. Which acceptance measure would reveal real progress?

  1. Measure end-to-end labor minutes per accepted outcome
  2. Measure risk-adjusted impact and implementation effort per use case
  3. Measure cost per accepted result against a defined business target
  4. Measure change in qualified escalations per thousand completed tasks

Explanation

Correct Answer

D:

The correct signal is change in qualified escalations per thousand completed tasks; it must be computed for representative workloads and compared with a baseline. It checks whether a measurable business outcome narrows the solution boundary before model selection while surfacing the exposure that feature demonstrations can obscure the absence of an accountable success condition.

Incorrect Answers

A:

The proposed measure would suit a setting where executives claim the assistant will save time without measuring current work, since measuring existing effort establishes a credible incremental-value estimate; however, that proxy would not demonstrate change in qualified escalations per thousand completed tasks for the selected intervention.

B:

The alternative indicator checks whether a scored decision framework surfaces high-value lower-risk starting points when candidate use cases differ in savings, uncertainty and personal-data exposure; but the acceptance gate must instead inspect change in qualified escalations per thousand completed tasks to detect the targeted failure.

C:

Where a team wants to fund a large platform before validating which use case helps, a relevant check would establish whether staged hypotheses tie each funding tranche to measured evidence; by contrast, the software platform release needs evidence from change in qualified escalations per thousand completed tasks.

Question 16

After installing Task-level acceptance rubric in an enterprise knowledge management workflow, the team must validate whether it has reduced the exposure that generic satisfaction scores may reward confident errors. Which measurement is most relevant?

  1. Audit high-percentile latency and timeout frequency
  2. Audit expert-reviewed acceptance rate by task category
  3. Audit verified factual claims divided by audited claims
  4. Audit total spend divided by validated resolved tickets

Explanation

Correct Answer

B:

A defensible evaluation looks at expert-reviewed acceptance rate by task category. The control's causal claim is that the evaluation process can define observable criteria for a usable output rather than aesthetic fluency. Monitoring that outcome, not merely elapsed tokens or success-looking text, helps catch generic satisfaction scores may reward confident errors.

Incorrect Answers

A:

Where interactive sessions frequently exceed their response deadline, a relevant check would establish whether the evaluation process can track p95 and p99 end-to-end latency by request class; but the acceptance gate must instead inspect expert-reviewed acceptance rate by task category to detect the targeted failure.

C:

When the system summarizes regulated policies with citations, the evaluation process can score whether factual claims are supported by cited source spans; however, that proxy would not demonstrate expert-reviewed acceptance rate by task category for the selected intervention.

D:

When model usage drops while human rework unexpectedly grows, the evaluation process can include review, retries and inference in outcome-based unit economics; but the acceptance gate must instead inspect expert-reviewed acceptance rate by task category to detect the targeted failure.

Question 17

A logistics coordination service owner is reviewing Workflow observation interviews after a canary rollout. The critical regression would occur if executive summaries alone omit edge cases that determine project success. What observable acceptance signal belongs in the dashboard?

  1. Track discovered exceptions covered by the agreed requirements
  2. Track proportion of nonfunctional requirements with acceptance tests
  3. Track baseline handoff count and exception cycle time
  4. Track agreed priority scores and expected benefits for shortlisted uses
  5. Track source coverage and freshness pass rate on sampled queries

Explanation

Correct Answer

A:

The release should evaluate discovered exceptions covered by the agreed requirements rather than equate successful calls with real outcomes. The control works on the premise that the delivery team can observe actual exception handling and frontline work before mapping automation; the proposed measure can expose the risk that executive summaries alone omit edge cases that determine project success.

Incorrect Answers

B:

When a proposal lists features but omits response time, recovery requirements and the delivery team can capture distinct behavior, latency, security, availability and data constraints; however, that proxy would not demonstrate discovered exceptions covered by the agreed requirements for the selected intervention.

C:

When automation is proposed without mapping present handoffs, failure paths and the delivery team can chart the existing input-to-outcome process and exception volumes; but the acceptance gate must instead inspect discovered exceptions covered by the agreed requirements to detect the targeted failure.

D:

The proposed measure would suit a setting where teams propose many attractive pilots without common selection criteria, since the delivery team can jointly rank candidates by value, feasibility, harm and data readiness; by contrast, the logistics coordination release needs evidence from discovered exceptions covered by the agreed requirements.

E:

When discovery assumes reliable source records that are fragmented or stale, the delivery team can validate quality, permissions, update cadence and ownership before promising RAG results; however, that proxy would not demonstrate discovered exceptions covered by the agreed requirements for the selected intervention.

Question 18

The telecommunications support release team needs a numerical gate for Ingress content classification. Rather than equating deployment with success, it wants to catch cases in which prompt-only guidance cannot neutralize malicious external text. Which indicator should it track?

  1. Evaluate attack containment time and residual failure severity
  2. Evaluate disallowed action attempts stopped at the executor
  3. Evaluate cross-tenant fairness and quota bypass incident counts
  4. Evaluate prompt-injection success rate against controlled attack fixtures

Explanation

Correct Answer

D:

An appropriate gate measures prompt-injection success rate against controlled attack fixtures. It tests the mechanism in which the control layer can classify untrusted inputs and prevent them from acquiring instruction authority, rather than a cosmetic demonstration. That distinction matters because prompt-only guidance cannot neutralize malicious external text.

Incorrect Answers

A:

Where failures may bypass one input or model safety layer, a relevant check would establish whether the control layer can combine prevention, detection, containment and audited recovery; however, that proxy would not demonstrate prompt-injection success rate against controlled attack fixtures for the selected intervention.

B:

The proposed measure would suit a setting where assistants propose privileged write operations in customer systems, since the control layer can require server-side authorization and business-policy checks before each side effect; but the acceptance gate must instead inspect prompt-injection success rate against controlled attack fixtures to detect the targeted failure.

C:

When one customer can monopolize shared model capacity through abusive retries, the control layer can enforce bounded usage budgets and anomaly controls per tenant. A favorable result on that test would not establish prompt-injection success rate against controlled attack fixtures for this workflow, where user-submitted attachments can contain instructions to steal private data; by contrast, the telecommunications support release needs evidence from prompt-injection success rate against controlled attack fixtures.

Question 19

A procurement services audit challenges the need for Task-scoped tool allowlist. Which risk would remain unmanaged if its function were not provided elsewhere?

  1. The danger that overlapping names lead to arbitrary selection and inconsistent answers
  2. The danger that broad tool visibility increases selection error and attack surface
  3. The danger that eager inclusion of the entire catalog wastes context and confuses selection
  4. The danger that uncontrolled retries exhaust quotas and conceal a failed precondition

Explanation

Correct Answer

B:

The loss of the control permits the situation where broad tool visibility increases selection error and attack surface. Its governing mechanism is that the integration can expose only tools needed for the active task and user permission scope; without an equivalent mechanism, a favorable demonstration cannot establish safety or quality.

Incorrect Answers

A:

When several plugins expose overlapping outdated customer lookup functions, overlapping names lead to arbitrary selection and inconsistent answers; whereas removing Task-scoped tool allowlist primarily exposes the risk that broad tool visibility increases selection error and attack surface.

C:

The competing hazard is that eager inclusion of the entire catalog wastes context and confuses selection in a workload where a complex connector ecosystem cannot fit full tool descriptions in each request; however, the named safeguard addresses the distinct hazard that broad tool visibility increases selection error and attack surface.

D:

If a search agent endlessly retries ineffective tools on malformed identifiers, uncontrolled retries exhaust quotas and conceal a failed precondition; whereas removing Task-scoped tool allowlist primarily exposes the risk that broad tool visibility increases selection error and attack surface.

Question 20

An operations analyst for enterprise knowledge management must evaluate Small-model first-pass routing on real tasks. A test must detect whether always using the strongest model increases cost and wait time, even when outputs appear fluent. Which measure is appropriate?

  1. Audit cost-quality-latency Pareto frontier on held-out examples
  2. Review audited synthesis accuracy on difficult cases
  3. Audit safe fallback completion rate under simulated outages
  4. Audit routed quality at the target cost and p95 latency

Explanation

Correct Answer

D:

Tracking routed quality at the target cost and p95 latency provides direct evidence that the system delivers its intended effect: a fast adequate tier handles easy cases and escalates only uncertain ones. The measure is more informative than prompt counts because the failure being measured is that always using the strongest model increases cost and wait time.

Incorrect Answers

A:

Where stakeholders choose a model based on brand rather than target workload, a relevant check would establish whether the model configuration can evaluate candidate models on identical task data and operating constraints; by contrast, the enterprise knowledge management release needs evidence from routed quality at the target cost and p95 latency.

B:

The proposed measure would suit a setting where multi-document arbitration requires several interdependent inferences, since a stronger model can justify tradeoffs across heterogeneous evidence; however, that proxy would not demonstrate routed quality at the target cost and p95 latency for the selected intervention.

C:

When primary inference service can time out during a critical support window, the model configuration can route to an approved fallback model with compatible output contracts; but the acceptance gate must instead inspect routed quality at the target cost and p95 latency to detect the targeted failure.

Claude Certified Architect - Professional (CCAR-P) Practice Tests