Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 78

Question 770

A healthcare operations pilot has enabled Privacy impact and lawful-basis review. Because technical encryption does not by itself establish regulatory permission to process, the release board wants evidence of effectiveness, not just a favorable demonstration. Which metric most directly tests the proposed control?

  1. Track forbidden region transfers found in data-flow audits
  2. Track documented lawful-basis coverage and purpose exceptions
  3. Track minimization compliance and utility impact by field
  4. Track vendor agreement coverage and PHI exposure incidents

Explanation

Correct Answer

B:

The release should evaluate documented lawful-basis coverage and purpose exceptions rather than equate successful calls with real outcomes. The control works on the premise that the control layer can involve privacy counsel in lawful basis, purpose and data minimization assessment; the proposed measure can expose the risk that technical encryption does not by itself establish regulatory permission to process.

Incorrect Answers

A:

The proposed measure would suit a setting where contract terms restrict where customer records may be processed or stored, since the control layer can verify regional processing, logs, backups and support access as a whole; by contrast, the healthcare operations release needs evidence from documented lawful-basis coverage and purpose exceptions.

C:

The proposed measure would suit a setting where classification only needs a narrow subset of full case records, since the control layer can transform and suppress unnecessary personal fields before invoking tools; but the acceptance gate must instead inspect documented lawful-basis coverage and purpose exceptions to detect the targeted failure.

D:

The alternative indicator checks whether the control layer can verify applicable HIPAA obligations and required business associate arrangements when a US clinical workflow handles protected health information through vendors; by contrast, the healthcare operations release needs evidence from documented lawful-basis coverage and purpose exceptions.

Question 771

An energy operations implementation team has approved Exact database query through a typed tool. Its expected behavior is that the integration can retrieve authorized structured records using parameterized queries. What prerequisite belongs on the release-readiness checklist?

  1. Prepare calibration dataset and a reviewed confidence threshold
  2. Prepare consistent snapshot semantics and dual access controls
  3. Prepare synonym inventory and a query expansion budget
  4. Prepare validated record IDs and least-privilege query access

Explanation

Correct Answer

D:

The required preparation is validated record IDs and least-privilege query access. It makes the following architecture claim enforceable: the integration can retrieve authorized structured records using parameterized queries. If that setup is skipped, generative similarity search is unsuitable for exact transactional state, so the release should examine exact-field match rate and permission denial correctness.

Incorrect Answers

A:

Where a knowledge search returns weak matches for an unsupported question, the integration can require evidence strength before generating a confident factual answer; for this release, the team must first verify validated record IDs and least-privilege query access.

B:

The setup is useful to enable the alternative's behavior where the integration can join exact database results with grounded policy excerpts in the application and financial answers need live numeric totals and current explanatory policy; however, the named intervention needs validated record IDs and least-privilege query access before its protection can be enforced.

C:

The setup is useful to enable the alternative's behavior where the integration can create bounded alternate queries and merge deduplicated matches and relevant policy is difficult to find because employees use many aliases; but the production dependency for Exact database query through a typed tool is validated record IDs and least-privilege query access.

Question 772

An independent assessment of a research administration Claude system reports that increasing temperature cannot restore clipped evidence. Which change best closes the exposed control gap?

  1. Enforce Model capability mismatch test
  2. Enforce Unsupported-claim provenance analysis
  3. Enforce Context truncation investigation
  4. Enforce Retrieval recall diagnosis

Explanation

Correct Answer

C:

The engineering fix changes the mechanism rather than the presentation: the evaluation process can inspect token budgets and the presence of decisive passages after packing. It requires serialized prompt capture with redaction; a regression test should examine missing-critical-span frequency versus context length in cases where increasing temperature cannot restore clipped evidence.

Incorrect Answers

A:

Although the alternative reduces the risk that deeper retrieval does not fix inadequate synthesis capability because the evaluation process can evaluate a more capable tier on the same evidence and instructions; yet it would not correct the observed weakness: increasing temperature cannot restore clipped evidence.

B:

This addresses the failure where adding more citation markers can disguise hallucinations, because the evaluation process can validate claim-to-span entailment rather than just link presence; but this remedy operates on another boundary, not the case where increasing temperature cannot restore clipped evidence.

D:

This addresses the failure where rewriting prompts cannot compensate for evidence never retrieved, because the evaluation process can inspect whether relevant source passages reached the model; yet it would not correct the observed weakness: increasing temperature cannot restore clipped evidence.

Question 773

Following a regression in telecommunications support, the team establishes that raw replay increases cost and may reintroduce superseded requests. Which proposed redesign most directly prevents recurrence?

  1. Prioritize Token-economics telemetry
  2. Prioritize Source-priority truncation policy
  3. Prioritize Rolling conversation summary with state
  4. Prioritize Exact token budget accounting

Explanation

Correct Answer

C:

The reported failure calls for the control under which the model configuration can preserve verified decisions and current state while pruning redundant history. A prompt-only reminder or unrelated metric would not address the risk that raw replay increases cost and may reintroduce superseded requests. Evidence of resolution requires explicit summary refresh and fact provenance and an improvement in carried-forward constraint accuracy across long conversations.

Incorrect Answers

A:

This addresses the failure where request totals alone cannot isolate token-heavy work, because the model configuration can record uncached input, cache reads, output and tool-related usage per task; but this remedy operates on another boundary, not the case where raw replay increases cost and may reintroduce superseded requests.

B:

This addresses the failure where first-in-first-out truncation may discard decisive exceptions, because the model configuration can truncate lowest-priority material while protecting essential instructions and citations; whereas the telecommunications support defect follows the path in which raw replay increases cost and may reintroduce superseded requests.

D:

When a character-count heuristic can understate model-specific token use, the model configuration can estimate input, tool-result and output allowances before sending requests; but this remedy operates on another boundary, not the case where raw replay increases cost and may reintroduce superseded requests.

Question 774

The media licensing release team needs a numerical gate for Shared artifact contracts. Rather than equating deployment with success, it wants to catch cases in which unconstrained natural-language handoffs lose provenance and state semantics. Which indicator should it track?

  1. Monitor end-to-end critical-path time and join completeness
  2. Monitor percentage of runs ending within declared budgets
  3. Monitor schema rejection rate and successful artifact joins
  4. Monitor rubric pass rate after a fixed maximum revision count

Explanation

Correct Answer

C:

The correct signal is schema rejection rate and successful artifact joins; it must be computed for representative workloads and compared with a baseline. It checks whether common schemas and immutable artifact versions make specialist outputs composable while surfacing the exposure that unconstrained natural-language handoffs lose provenance and state semantics.

Incorrect Answers

A:

The alternative indicator checks whether the architecture can run independent checks concurrently and combine results with deterministic joins when multiple independent document checks dominate total processing time; however, that proxy would not demonstrate schema rejection rate and successful artifact joins for the selected intervention.

B:

Where a planner sometimes delegates recursively until requests time out, a relevant check would establish whether the architecture can impose step, token, cost and time ceilings with explicit exit behavior; but the acceptance gate must instead inspect schema rejection rate and successful artifact joins to detect the targeted failure.

D:

The alternative indicator checks whether a separate evaluation pass checks defined criteria and requests capped revisions when drafted release notes repeatedly miss mandatory restrictions; however, that proxy would not demonstrate schema rejection rate and successful artifact joins for the selected intervention.

Question 775

The post-incident analysis for legal operations concludes that a high-level component box diagram can hide secondary retention paths. Which remediation targets the relevant failure boundary?

  1. Introduce End-to-end data-flow diagram
  2. Introduce Prompt and model configuration manifest
  3. Trace each architecture decision to its originating requirement
  4. Introduce Deployment topology and residency record

Explanation

Correct Answer

A:

The problem lies in the absence of a mechanism where the delivery team can show all data stores, transformations, transfers and trust boundaries. Reinstating it, backed by service inventory and data classification, limits the exposure that a high-level component box diagram can hide secondary retention paths; the engineering team should track completeness of identified data-flow and access boundaries.

Incorrect Answers

B:

Where multiple environments produce divergent behavior despite using one codebase, the delivery team can version prompts, model choices, tools and safety settings together; but this remedy operates on another boundary, not the case where a high-level component box diagram can hide secondary retention paths.

C:

This addresses the failure where documentation volume without traceability is hard to audit or change, because the delivery team can link decisions to acceptance tests and authoritative requirement IDs; whereas the legal operations defect follows the path in which a high-level component box diagram can hide secondary retention paths.

D:

When an unqualified reference diagram cannot prove actual deployment controls, the delivery team can document region, identity, storage and network paths for each environment; yet it would not correct the observed weakness: a high-level component box diagram can hide secondary retention paths.

Question 776

A logistics coordination release candidate fails review after assessors observe that technical encryption does not by itself establish regulatory permission to process. What intervention should take priority?

  1. Deploy Retention and erasure workflow
  2. Deploy Regulatory evidence package
  3. Deploy HIPAA responsibility and contract assessment
  4. Deploy Privacy impact and lawful-basis review

Explanation

Correct Answer

D:

The remediation acts at the relevant design boundary: the control layer can involve privacy counsel in lawful basis, purpose and data minimization assessment. That counteracts the observed hazard that technical encryption does not by itself establish regulatory permission to process. Rollout should include jurisdiction mapping and data-use inventory, with regression evidence from documented lawful-basis coverage and purpose exceptions.

Incorrect Answers

A:

Although the alternative reduces the risk that deleting a front-end record may leave copies in downstream storage because the control layer can map and process erasure across prompts, logs, caches and derived indexes as applicable; but this remedy operates on another boundary, not the case where technical encryption does not by itself establish regulatory permission to process.

B:

When an informal assurance statement cannot reconstruct a production event, the control layer can keep traceable control evidence, access decisions and versioned approvals; whereas the logistics coordination defect follows the path in which technical encryption does not by itself establish regulatory permission to process.

C:

Where a US clinical workflow handles protected health information through vendors, the control layer can verify applicable HIPAA obligations and required business associate arrangements; yet it would not correct the observed weakness: technical encryption does not by itself establish regulatory permission to process.

Question 777

An enterprise knowledge management pilot has enabled Exact database query through a typed tool. Because generative similarity search is unsuitable for exact transactional state, the release board wants evidence of effectiveness, not just a favorable demonstration. Which metric most directly tests the proposed control?

  1. Monitor composite answer correctness for numbers and policy conditions
  2. Monitor exact-field match rate and permission denial correctness
  3. Monitor version-correct answer rate on dated cases
  4. Monitor path validity and provenance of returned edges

Explanation

Correct Answer

B:

An appropriate gate measures exact-field match rate and permission denial correctness. It tests the mechanism in which the integration can retrieve authorized structured records using parameterized queries, rather than a cosmetic demonstration. That distinction matters because generative similarity search is unsuitable for exact transactional state.

Incorrect Answers

A:

The proposed measure would suit a setting where financial answers need live numeric totals and current explanatory policy, since the integration can join exact database results with grounded policy excerpts in the application; however, that proxy would not demonstrate exact-field match rate and permission denial correctness for the selected intervention.

C:

Where users ask what rule applied on a particular historical date, a relevant check would establish whether the integration can filter versions by effective interval before selecting passages; by contrast, the enterprise knowledge management release needs evidence from exact-field match rate and permission denial correctness.

D:

When investigators ask how two entities connect through approved relationships, the integration can traverse an authorized graph rather than rely on plain passage similarity; however, that proxy would not demonstrate exact-field match rate and permission denial correctness for the selected intervention.

Question 778

The energy operations architect proposes omitting Context truncation investigation from the production design. What is the most direct failure mode that decision would expose?

  1. A failure where adding more citation markers can disguise hallucinations
  2. A failure where increasing temperature cannot restore clipped evidence
  3. A failure where fluent restatement can hide intermediate fabrication
  4. A failure where rewriting prompts cannot compensate for evidence never retrieved

Explanation

Correct Answer

B:

The loss of the control permits the situation where increasing temperature cannot restore clipped evidence. Its governing mechanism is that the evaluation process can inspect token budgets and the presence of decisive passages after packing; without an equivalent mechanism, a favorable demonstration cannot establish safety or quality.

Incorrect Answers

A:

If a response confidently cites text that does not support the claim, adding more citation markers can disguise hallucinations; whereas removing Context truncation investigation primarily exposes the risk that increasing temperature cannot restore clipped evidence.

C:

The competing hazard is that fluent restatement can hide intermediate fabrication in a workload where an agent selects correct tools but invents values absent from responses; however, the named safeguard addresses the distinct hazard that increasing temperature cannot restore clipped evidence.

D:

If generated answers omit a rule even though it exists in the knowledge base, rewriting prompts cannot compensate for evidence never retrieved; whereas removing Context truncation investigation primarily exposes the risk that increasing temperature cannot restore clipped evidence.

Question 779

A lean redesign in logistics coordination would retire Parallel code-review assistance while leaving the rest of the service unchanged. Which problem would remain without the retired capability?

  1. The danger that a single sweeping rewrite is hard to validate and roll back
  2. The danger that one general review prompt may miss specialist concerns
  3. The danger that unattributed code changes weaken auditability and incident diagnosis
  4. The danger that unverified generated patches can introduce silent regressions

Explanation

Correct Answer

B:

The relevant failure mode is that one general review prompt may miss specialist concerns. The existing design can prevent it when the engineering workflow can run security, maintainability and API compatibility reviews separately. The team needs fixed review rubrics and deduplicated findings and must monitor distinct actionable findings validated by maintainers to support any claim of equivalent protection.

Incorrect Answers

A:

The competing hazard is that a single sweeping rewrite is hard to validate and roll back in a workload where a large codebase migration would touch many coupled modules; but this is not the immediate loss associated with omitting Parallel code-review assistance: one general review prompt may miss specialist concerns.

C:

If release teams cannot tell why generated modifications were proposed, unattributed code changes weaken auditability and incident diagnosis; whereas removing Parallel code-review assistance primarily exposes the risk that one general review prompt may miss specialist concerns.

D:

The competing hazard is that unverified generated patches can introduce silent regressions in a workload where agents generate plausible code that breaks existing behavior; but this is not the immediate loss associated with omitting Parallel code-review assistance: one general review prompt may miss specialist concerns.

Question 780

Following a regression in healthcare operations, the team establishes that unconstrained natural-language handoffs lose provenance and state semantics. Which proposed redesign most directly prevents recurrence?

  1. Introduce Supervisor-mediated delegation
  2. Introduce Independent critic with bounded revisions
  3. Introduce Parallel independent subtask fan-out
  4. Introduce Sequential dependency-aware handoffs
  5. Introduce Shared artifact contracts

Explanation

Correct Answer

E:

This is the causal correction: common schemas and immutable artifact versions make specialist outputs composable. With versioned data contracts and ownership of identifiers in place, the team can test schema rejection rate and successful artifact joins and show that the problematic case in which unconstrained natural-language handoffs lose provenance and state semantics has been addressed.

Incorrect Answers

A:

Where several specialists must cooperate without granting every one all tools, a supervisor assigns tasks and consolidates results within distinct permission scopes; whereas the healthcare operations defect follows the path in which unconstrained natural-language handoffs lose provenance and state semantics.

B:

This addresses the failure where unbounded self-critique loops can raise cost without a convergence rule, because a separate evaluation pass checks defined criteria and requests capped revisions; yet it would not correct the observed weakness: unconstrained natural-language handoffs lose provenance and state semantics.

C:

This addresses the failure where serial execution creates avoidable critical-path latency, because the architecture can run independent checks concurrently and combine results with deterministic joins; but this remedy operates on another boundary, not the case where unconstrained natural-language handoffs lose provenance and state semantics.

D:

The alternative reduces the risk that parallelizing dependent tasks can validate stale or missing artifacts because the architecture can orchestrate a staged dependency graph with explicit prerequisites; however, the healthcare operations defect follows the path in which unconstrained natural-language handoffs lose provenance and state semantics.

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