Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 17

Question 160

An operations analyst for insurance processing must evaluate Use-case prioritization workshop on real tasks. A test must detect whether political priority alone may select a low-return high-risk deployment, even when outputs appear fluent. Which measure is appropriate?

  1. Audit agreed priority scores and expected benefits for shortlisted uses
  2. Audit proportion of nonfunctional requirements with acceptance tests
  3. Audit outstanding decision-owner ambiguities before design signoff
  4. Audit discovered exceptions covered by the agreed requirements

Explanation

Correct Answer

A:

The release should evaluate agreed priority scores and expected benefits for shortlisted uses rather than equate successful calls with real outcomes. The control works on the premise that the delivery team can jointly rank candidates by value, feasibility, harm and data readiness; the proposed measure can expose the risk that political priority alone may select a low-return high-risk deployment.

Incorrect Answers

B:

Where a proposal lists features but omits response time and recovery requirements, a relevant check would establish whether the delivery team can capture distinct behavior, latency, security, availability and data constraints; by contrast, the insurance processing release needs evidence from agreed priority scores and expected benefits for shortlisted uses.

C:

When product, security, legal teams disagree on who can approve release and the delivery team can define accountable decision owners and consultation rights early. A favorable result on that test would not establish agreed priority scores and expected benefits for shortlisted uses for this workflow, where; however, that proxy would not demonstrate agreed priority scores and expected benefits for shortlisted uses for the selected intervention.

D:

Where sponsors describe an ideal process that differs from daily operations, a relevant check would establish whether the delivery team can observe actual exception handling and frontline work before mapping automation; but the acceptance gate must instead inspect agreed priority scores and expected benefits for shortlisted uses to detect the targeted failure.

Question 161

During technical sign-off for legal operations, Per-user delegated authorization has been selected. The design premise is that the integration can enforce each call using the requesting user's authorized scope. Which enabling condition must be demonstrated before production traffic is allowed?

  1. Arrange signing keys and bounded timestamp tolerance
  2. Arrange authoritative tenant IDs and indexing ACL metadata
  3. Arrange trusted auth context and policy engine
  4. Arrange propagated identity and resource-level access policy

Explanation

Correct Answer

D:

The required preparation is propagated identity and resource-level access policy. It makes the following architecture claim enforceable: the integration can enforce each call using the requesting user's authorized scope. If that setup is skipped, a shared service token can reveal records across users, so the release should examine cross-tenant unauthorized-read attempts successfully denied.

Incorrect Answers

A:

For the separate challenge where an upstream application posts actions to a model-driven workflow, the integration can verify message authenticity, timestamp and replay protection before accepting commands; but the production dependency for Per-user delegated authorization is propagated identity and resource-level access policy.

B:

Where one vector index holds embeddings for several client organizations, the integration can apply tenant and document ACL filters before results reach the model; for this release, the team must first verify propagated identity and resource-level access policy.

C:

Where a model can propose actions but must not decide its own privileges, the executor evaluates policy before invoking protected downstream APIs; however, the named intervention needs propagated identity and resource-level access policy before its protection can be enforced.

Question 162

The logistics coordination deployment uses High-risk human approval gate, but executive reports currently count only successful model responses. The residual concern is that after-the-fact logs cannot reverse a completed sensitive transfer. Which acceptance measure would reveal real progress?

  1. Audit prompt-injection success rate against controlled attack fixtures
  2. Audit unapproved high-risk actions prevented during simulations
  3. Audit cross-tenant fairness and quota bypass incident counts
  4. Audit policy override rate in hostile-document evaluations
  5. Audit disallowed action attempts stopped at the executor

Explanation

Correct Answer

B:

The acceptance measure focuses on unapproved high-risk actions prevented during simulations at the task or workflow boundary. Model fluency would not establish whether the control layer can stop before execution and require authorized human confirmation, and it could conceal cases where after-the-fact logs cannot reverse a completed sensitive transfer.

Incorrect Answers

A:

The alternative indicator checks whether the control layer can classify untrusted inputs and prevent them from acquiring instruction authority when user-submitted attachments can contain instructions to steal private data; however, that proxy would not demonstrate unapproved high-risk actions prevented during simulations for the selected intervention.

C:

The alternative indicator checks whether the control layer can enforce bounded usage budgets and anomaly controls per tenant when one customer can monopolize shared model capacity through abusive retries; by contrast, the logistics coordination release needs evidence from unapproved high-risk actions prevented during simulations.

D:

Where supplier documents include embedded instructions masquerading as system messages, a relevant check would establish whether the control layer can treat retrieved text as evidence and prohibit it from changing tool policy; however, that proxy would not demonstrate unapproved high-risk actions prevented during simulations for the selected intervention.

E:

Where assistants propose privileged write operations in customer systems, a relevant check would establish whether the control layer can require server-side authorization and business-policy checks before each side effect; but the acceptance gate must instead inspect unapproved high-risk actions prevented during simulations to detect the targeted failure.

Question 163

After installing Phased value hypothesis testing in a human-resources services workflow, the team must validate whether it has reduced the exposure that a portfolio of speculative features can crowd out working high-value uses. Which measurement is most relevant?

  1. Measure end-to-end labor minutes per accepted outcome
  2. Measure pass rate on a stratified set of authentic exceptions
  3. Measure reviewer override and acceptance rates by decision category
  4. Measure cost per accepted result against a defined business target
  5. Measure change in qualified escalations per thousand completed tasks

Explanation

Correct Answer

D:

Tracking cost per accepted result against a defined business target provides direct evidence that the system delivers its intended effect: staged hypotheses tie each funding tranche to measured evidence. The measure is more informative than prompt counts because the failure being measured is that a portfolio of speculative features can crowd out working high-value uses.

Incorrect Answers

A:

Where executives claim the assistant will save time without measuring current work, a relevant check would establish whether measuring existing effort establishes a credible incremental-value estimate; but the acceptance gate must instead inspect cost per accepted result against a defined business target to detect the targeted failure.

B:

The alternative indicator checks whether small-scope trials expose feasibility and exception handling before a broad commitment when the proposed automation encounters exceptions that have never been tested; by contrast, the human-resources services release needs evidence from cost per accepted result against a defined business target.

C:

When caseworkers must retain authority for consequential final decisions, Claude drafts ranked recommendations while a designated reviewer authorizes each action; however, that proxy would not demonstrate cost per accepted result against a defined business target for the selected intervention.

E:

Where stakeholders disagree whether faster drafting or fewer escalations is the goal, a relevant check would establish whether a measurable business outcome narrows the solution boundary before model selection; by contrast, the human-resources services release needs evidence from cost per accepted result against a defined business target.

Question 164

A travel operations team is advancing Reproducible development environment beyond its proof of concept. Engineers expect that the engineering workflow can pin dependencies, tooling and environment assumptions across contributors. What evidence of setup readiness should be required first?

  1. Establish central connector catalog and credential lifecycle
  2. Establish approved project guidance and code review owners
  3. Establish versioned setup scripts and build reproducibility checks
  4. Establish isolated identities and explicit workspace inventory

Explanation

Correct Answer

C:

The proposed design needs versioned setup scripts and build reproducibility checks as its foundation. This matters because the engineering workflow can pin dependencies, tooling and environment assumptions across contributors; a plausible demonstration without the prerequisite can still leave a case where ambient local settings conceal incompatibilities in shared builds. The relevant check is agent-generated changes passing clean-environment builds.

Incorrect Answers

A:

This prerequisite supports a method used when development teams connect tools with overlapping credentials and contracts, because the engineering workflow can register reviewed MCP servers with clear ownership and allowed scopes; but the production dependency for Reproducible development environment is versioned setup scripts and build reproducibility checks.

B:

This prerequisite supports a method used when developers get inconsistent coding behavior across branches and members, because the engineering workflow can version shared Claude Code guidance alongside the project and review changes; for this release, the team must first verify versioned setup scripts and build reproducibility checks.

D:

This prerequisite supports a method used when confidential projects share agent context and local tool access inadvertently, because the engineering workflow can isolate project roots, secrets and tool permissions by approved workspace; but the production dependency for Reproducible development environment is versioned setup scripts and build reproducibility checks.

Question 165

For a software platform rollout, the proposed design uses Context-capacity fit before model selection, relying on the claim that the model configuration can test effective context utilization and recall before choosing a model configuration. What enabling control must be in place to make that design enforceable?

  1. Establish representative examples and equalized test settings
  2. Establish benchmark cases with multi-hop evidence
  3. Establish long-context benchmark cases and token accounting
  4. Establish model contract compatibility tests and outage policy

Explanation

Correct Answer

C:

Readiness depends on long-context benchmark cases and token accounting, not just a design approval. That condition supports the mechanism in which the model configuration can test effective context utilization and recall before choosing a model configuration. Without it, nominal context limits alone do not prove reliable use of every passage; monitoring relevant-fact recall by document position and input length helps verify the deployed result.

Incorrect Answers

A:

For the separate challenge 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; for this release, the team must first verify long-context benchmark cases and token accounting.

B:

This prerequisite supports a method used when multi-document arbitration requires several interdependent inferences, because a stronger model can justify tradeoffs across heterogeneous evidence; however, the named intervention needs long-context benchmark cases and token accounting before its protection can be enforced.

D:

For the separate challenge where 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; for this release, the team must first verify long-context benchmark cases and token accounting.

Question 166

The post-incident analysis for procurement services concludes that a successful business task may still leak restricted records. Which remediation targets the relevant failure boundary?

  1. Enforce Grounded-claim verification rate
  2. Enforce Task-level acceptance rubric
  3. Enforce Security policy violation rate
  4. Enforce Calibration and abstention quality

Explanation

Correct Answer

C:

This intervention addresses the failure at its source because the evaluation process can evaluate denied and erroneously permitted requests against authoritative policy. To keep the issue from recurring, establish labeled authorization test cases and verify false-allow and false-deny rates by scope; otherwise a successful business task may still leak restricted records can persist.

Incorrect Answers

A:

Where the system summarizes regulated policies with citations, the evaluation process can score whether factual claims are supported by cited source spans; yet it would not correct the observed weakness: a successful business task may still leak restricted records.

B:

This addresses the failure where generic satisfaction scores may reward confident errors, because the evaluation process can define observable criteria for a usable output rather than aesthetic fluency; but this remedy operates on another boundary, not the case where a successful business task may still leak restricted records.

D:

Where incomplete evidence should yield escalation rather than invented certainty, the evaluation process can evaluate correctness jointly with justified abstentions across uncertainty levels; yet it would not correct the observed weakness: a successful business task may still leak restricted records.

Question 167

An operations analyst for financial operations must evaluate Per-user delegated authorization on real tasks. A test must detect whether a shared service token can reveal records across users, even when outputs appear fluent. Which measure is appropriate?

  1. Audit credential scope violations and secret age
  2. Audit revoked-token access denial and scope mismatch rate
  3. Audit cross-tenant unauthorized-read attempts successfully denied
  4. Audit cross-tenant retrieval violations in isolation tests
  5. Audit blocked unauthorized operations in adversarial tool-call tests

Explanation

Correct Answer

C:

The acceptance measure focuses on cross-tenant unauthorized-read attempts successfully denied at the task or workflow boundary. Model fluency would not establish whether the integration can enforce each call using the requesting user's authorized scope, and it could conceal cases where a shared service token can reveal records across users.

Incorrect Answers

A:

Where unattended background jobs need limited read access to a narrow dataset, a relevant check would establish whether the integration can issue narrowly scoped renewable credentials for the job identity; by contrast, the financial operations release needs evidence from cross-tenant unauthorized-read attempts successfully denied.

B:

The proposed measure would suit a setting where employees connect personal enterprise resources through an integration, since the integration can obtain user consent for minimum scopes and handle revocation; however, that proxy would not demonstrate cross-tenant unauthorized-read attempts successfully denied for the selected intervention.

D:

Where one vector index holds embeddings for several client organizations, a relevant check would establish whether the integration can apply tenant and document ACL filters before results reach the model; by contrast, the financial operations release needs evidence from cross-tenant unauthorized-read attempts successfully denied.

E:

When a model can propose actions but must not decide its own privileges, the executor evaluates policy before invoking protected downstream APIs; however, that proxy would not demonstrate cross-tenant unauthorized-read attempts successfully denied for the selected intervention.

Question 168

A supply-chain planning incident exposes a specific weakness: a portfolio of speculative features can crowd out working high-value uses. Which architectural correction addresses the cause rather than masking the symptom?

  1. Enforce Human-approved recommendation service
  2. Enforce Outcome-first use-case framing
  3. Enforce Feasibility pilot with representative cases
  4. Enforce Phased value hypothesis testing

Explanation

Correct Answer

D:

The reported failure calls for the control under which staged hypotheses tie each funding tranche to measured evidence. A prompt-only reminder or unrelated metric would not address the risk that a portfolio of speculative features can crowd out working high-value uses. Evidence of resolution requires hypotheses with owners and stage-gate thresholds and an improvement in cost per accepted result against a defined business target.

Incorrect Answers

A:

When automating the last step could bypass required accountability, Claude drafts ranked recommendations while a designated reviewer authorizes each action; whereas the supply-chain planning defect follows the path in which a portfolio of speculative features can crowd out working high-value uses.

B:

When feature demonstrations can obscure the absence of an accountable success condition, a measurable business outcome narrows the solution boundary before model selection; yet it would not correct the observed weakness: a portfolio of speculative features can crowd out working high-value uses.

C:

Where the proposed automation encounters exceptions that have never been tested, small-scope trials expose feasibility and exception handling before a broad commitment; but this remedy operates on another boundary, not the case where a portfolio of speculative features can crowd out working high-value uses.

Question 169

The healthcare operations staging review finds that political priority alone may select a low-return high-risk deployment. Which design intervention is the most appropriate corrective action?

  1. Deploy Use-case prioritization workshop
  2. Deploy Workflow observation interviews
  3. Deploy Functional versus nonfunctional split
  4. Deploy Data readiness assessment
  5. Deploy Current-state service blueprint
  6. Deploy Stakeholder and decision-rights mapping

Explanation

Correct Answer

A:

This is the causal correction: the delivery team can jointly rank candidates by value, feasibility, harm and data readiness. With scoring rubric and representatives from delivery and risk teams in place, the team can test agreed priority scores and expected benefits for shortlisted uses and show that the problematic case in which political priority alone may select a low-return high-risk deployment has been addressed.

Incorrect Answers

B:

Although the alternative reduces the risk that executive summaries alone omit edge cases that determine project success because the delivery team can observe actual exception handling and frontline work before mapping automation; but this remedy operates on another boundary, not the case where political priority alone may select a low-return high-risk deployment.

C:

This addresses the failure where feature completeness does not demonstrate production acceptability, because the delivery team can capture distinct behavior, latency, security, availability and data constraints; whereas the healthcare operations defect follows the path in which political priority alone may select a low-return high-risk deployment.

D:

This addresses the failure where a strong model cannot compensate for absent or unauthorized source data, because the delivery team can validate quality, permissions, update cadence and ownership before promising RAG results; yet it would not correct the observed weakness: political priority alone may select a low-return high-risk deployment.

E:

When redesigning a guessed process can shift work rather than remove it, the delivery team can chart the existing input-to-outcome process and exception volumes; but this remedy operates on another boundary, not the case where political priority alone may select a low-return high-risk deployment.

F:

Where product, security and legal teams disagree on who can approve release, the delivery team can define accountable decision owners and consultation rights early; whereas the healthcare operations defect follows the path in which political priority alone may select a low-return high-risk deployment.

Question 170

An independent assessment of a telecommunications support Claude system reports that after-the-fact logs cannot reverse a completed sensitive transfer. Which change best closes the exposed control gap?

  1. Apply Defense-in-depth incident response
  2. Apply High-risk human approval gate
  3. Apply Output privacy filter
  4. Apply Rate and spend limits by tenant

Explanation

Correct Answer

B:

The problem lies in the absence of a mechanism where the control layer can stop before execution and require authorized human confirmation. Reinstating it, backed by reviewer authority, timeout path and approval evidence, limits the exposure that after-the-fact logs cannot reverse a completed sensitive transfer; the engineering team should track unapproved high-risk actions prevented during simulations.

Incorrect Answers

A:

This addresses the failure where a single classifier cannot handle every emergent attack pattern, because the control layer can combine prevention, detection, containment and audited recovery; but this remedy operates on another boundary, not the case where after-the-fact logs cannot reverse a completed sensitive transfer.

C:

This addresses the failure where upstream input minimization alone may not catch derived disclosures, because the control layer can enforce recipient-specific redaction and validated output fields; yet it would not correct the observed weakness: after-the-fact logs cannot reverse a completed sensitive transfer.

D:

Where one customer can monopolize shared model capacity through abusive retries, the control layer can enforce bounded usage budgets and anomaly controls per tenant; but this remedy operates on another boundary, not the case where after-the-fact logs cannot reverse a completed sensitive transfer.

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