Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 37

Question 360

A regional deployment uses a different approval workflow but no record explains the exception. What should the solution architect prioritize?

  1. Maintain a decision log tying permissions to reviewers and specific workflow justifications.
  2. Document localization decisions and the legal/control basis for each variation.
  3. Document SLIs, alert thresholds, owners and operational playbooks.
  4. Record versioned dependencies and configuration in a deployment manifest.

Explanation

Correct Answer

B:

Operational deviations need context so maintainers do not mistakenly remove necessary safeguards. Version the artifacts proving governed regional configuration differences with associated configuration and verification records so a successor can reproduce the decision.

Incorrect Answers

A:

It emphasizes auditable privilege approval history without closing the stated gap in governed regional configuration differences. By contrast, the remedy to document localization decisions and the legal/control basis for each variation acts on the condition that makes this scenario fail.

C:

Improving service reliability handoff completeness does not imply governed regional configuration differences. The selected design must document localization decisions and the legal/control basis for each variation, whereas this choice omits that direct protection or verification step.

D:

This recommendation targets complete deployment configuration provenance at a different point in the workflow. It leaves governed regional configuration differences outside the design contract and does not require the team to document localization decisions and the legal/control basis for each variation.

Question 361

A team changes the prompt policy and forgets to update its evaluation rubric. Which design choice best matches the stated constraint?

  1. Record architectural assumptions, verified constraints and exceptions in a decision record.
  2. Document training/adaptation provenance, evaluation scope and known limitations.
  3. Document localization decisions and the legal/control basis for each variation.
  4. Keep prompt, policy, test and release notes linked through change control.

Explanation

Correct Answer

D:

The validity of test evidence depends on the behavior and constraints actually deployed. Version the artifacts proving alignment of documentation and verification with associated configuration and verification records so a successor can reproduce the decision.

Incorrect Answers

A:

The relevant trade-off is not primarily explicit reusable architecture assumptions. A successful fix needs alignment of documentation and verification; otherwise alignment of documentation and verification remains unverified even after this change.

B:

The control would be justified if transparent customized-model provenance were the limiting factor. In this scenario the decisive target is alignment of documentation and verification; success requires an architecture that will keep prompt, policy, test and release notes linked through change control.

C:

The option belongs to the broader work on governed regional configuration differences, but it cannot by itself deliver alignment of documentation and verification. A decision on this case needs the team to keep prompt, policy, test and release notes linked through change control and verify that outcome.

Question 362

A high-stakes pilot has no named on-call owner, alerting or usable rollback gate. Its production connector also depends on credentials that have never been rotated under the service identity used after launch. Which TWO readiness activities are essential? Choose TWO.

  1. Implement a governed complaint and trace-reconstruction process.
  2. Block or roll back release based on segment-specific nonnegotiable thresholds.
  3. Name an on-call owner in launch notes without adding tested alerts, incident routing or rollback procedures.
  4. Add operational acceptance gates before calling it production-ready.
  5. Rehearse connector credential rotation and reauthentication under the actual production service identity.

Explanation

Correct Answers

D:

A strong sandbox demonstration is not enough for a high-stakes production handoff. On-call ownership, alert thresholds, abort conditions and rollback rehearsals must be acceptance gates; a live-identity secret-rotation test separately proves connector readiness.

E:

A test run with developer credentials does not prove that production integrations will survive credential expiry or rotation. An authorized rotation rehearsal exposes gaps in secret distribution, permissions and reconnect recovery.

Incorrect Answers

A:

The design concern is real but different: audit-supported customer remediation. It does not impose the specific requirement for production readiness beyond proof of concept and production-identity credential rotation rehearsal; a suitable solution must add operational acceptance gates before calling it production-ready and demonstrate the result.

B:

This addresses risk-gated lifecycle evolution; however, the problem asks for production readiness beyond proof of concept and production-identity credential rotation rehearsal. Without the safeguard to rehearse connector credential rotation and reauthentication under the actual production service identity, the original failure mode is still possible.

C:

Nominal assignment does not make a service operationally ready. The rollout requires exercised operational gates and a separate connector rotation test under its production identity.

Question 363

A new AI workflow launches nationally without staged traffic or a recovery plan. Which next step is most defensible?

  1. Block or roll back release based on segment-specific nonnegotiable thresholds.
  2. Execute decommissioning that revokes credentials and applies retention/deletion rules.
  3. Use canary rollout, predeclared abort thresholds and tested rollback.
  4. Complete staged security and entitlement testing under real roles before expanding access.

Explanation

Correct Answer

C:

Progressive release limits blast radius and validates actual conditions. Assign operational ownership of risk-managed deployment lifecycle and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

This recommendation targets risk-gated lifecycle evolution at a different point in the workflow. It leaves risk-managed deployment lifecycle outside the design contract and does not require the team to use canary rollout, predeclared abort thresholds and tested rollback.

B:

This is a valid architectural concern when optimizing secure agent decommissioning. Here it leaves risk-managed deployment lifecycle unaddressed, whereas the intended change must use canary rollout, predeclared abort thresholds and tested rollback before the case can be considered resolved.

D:

By itself, this improves real-identity prelaunch verification, but the case is not asking the architect to optimize that dimension. The requirement is risk-managed deployment lifecycle, supported by an approach to use canary rollout, predeclared abort thresholds and tested rollback.

Question 364

A legal policy changes and the assistant keeps using the old retrieval index for several weeks. Which approach should be recommended?

  1. Use canary rollout, predeclared abort thresholds and tested rollback.
  2. Execute decommissioning that revokes credentials and applies retention/deletion rules.
  3. Create change ownership and refresh procedures tied to source updates.
  4. Block or roll back release based on segment-specific nonnegotiable thresholds.
  5. Assign operating owners, on-call escalation and maintenance budgets.

Explanation

Correct Answer

C:

The content lifecycle must be maintained after initial model deployment. Assign operational ownership of continuous knowledge currency and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

This shifts attention toward risk-managed deployment lifecycle when the relevant decision is about continuous knowledge currency. Unless the architecture can create change ownership and refresh procedures tied to source updates, the failure described in the stem remains unresolved.

B:

For a workflow whose primary weakness was secure agent decommissioning, that proposal could be reasonable. Here the unmet obligation is continuous knowledge currency, and that requires the system to create change ownership and refresh procedures tied to source updates.

D:

It is aimed at a different outcome, risk-gated lifecycle evolution. No evidence is given that it provides continuous knowledge currency; the needed mechanism is to create change ownership and refresh procedures tied to source updates at the appropriate boundary.

E:

It emphasizes post-launch operational ownership without closing the stated gap in continuous knowledge currency. By contrast, the remedy to create change ownership and refresh procedures tied to source updates acts on the condition that makes this scenario fail.

Question 365

A deployment team's ownership ends on launch day even though ongoing incident response is required. Which change most directly satisfies this requirement?

  1. Use prioritized intake categories with owners, severity and response SLAs.
  2. Create change ownership and refresh procedures tied to source updates.
  3. Assign operating owners, on-call escalation and maintenance budgets.
  4. Add operational acceptance gates before calling it production-ready.

Explanation

Correct Answer

C:

Production systems need persistent accountability across the whole lifecycle. Assign operational ownership of post-launch operational ownership and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

There is a legitimate use case for structured post-launch feedback triage, but it does not supply the test or enforcement needed for post-launch operational ownership. The design needs to assign operating owners, on-call escalation and maintenance budgets.

B:

Consider its actual operating scope: continuous knowledge currency. The requested assurance is post-launch operational ownership, which calls for an implementation that can assign operating owners, on-call escalation and maintenance budgets, not merely for a related improvement.

D:

The proposed measure is aimed at production readiness beyond proof of concept, not post-launch operational ownership. The architecture must instead assign operating owners, on-call escalation and maintenance budgets; a separate change is still needed for post-launch operational ownership.

Question 366

A quality team learns of recurrent exceptions but has no process to feed them into evaluation data. Which solution best resolves this operational gap?

  1. Block or roll back release based on segment-specific nonnegotiable thresholds.
  2. Use prioritized intake categories with owners, severity and response SLAs.
  3. Triage incidents into versioned regression cases with prioritized remediation.
  4. Implement a governed complaint and trace-reconstruction process.

Explanation

Correct Answer

C:

Real failures should expand the test suite and inform subsequent releases. Assign operational ownership of incident-driven evaluation improvement and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

By itself, this improves risk-gated lifecycle evolution, but the case is not asking the architect to optimize that dimension. The requirement is incident-driven evaluation improvement, supported by an approach to triage incidents into versioned regression cases with prioritized remediation.

B:

Its main effect concerns structured post-launch feedback triage. The case explicitly hinges on incident-driven evaluation improvement; to resolve it reliably, the implementation has to triage incidents into versioned regression cases with prioritized remediation, which this choice does not establish.

D:

The proposed action prioritizes audit-supported customer remediation. That can coexist with the right solution, but it cannot replace incident-driven evaluation improvement: the application must still triage incidents into versioned regression cases with prioritized remediation.

Question 367

A high-risk assistant passes sandbox tests but access controls are not verified with actual staff accounts. What design decision would best fit these facts?

  1. Add operational acceptance gates before calling it production-ready.
  2. Complete staged security and entitlement testing under real roles before expanding access.
  3. Block or roll back release based on segment-specific nonnegotiable thresholds.
  4. Use canary rollout, predeclared abort thresholds and tested rollback.

Explanation

Correct Answer

B:

Production readiness requires identity and data-path validation outside developer accounts. Assign operational ownership of real-identity prelaunch verification and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

It is aimed at a different outcome, production readiness beyond proof of concept. No evidence is given that it provides real-identity prelaunch verification; the needed mechanism is to complete staged security and entitlement testing under real roles before expanding access at the appropriate boundary.

C:

That is primarily a technique for risk-gated lifecycle evolution. It lacks the explicit measure for real-identity prelaunch verification that is necessary here, namely a system that will complete staged security and entitlement testing under real roles before expanding access.

D:

Improving risk-managed deployment lifecycle does not imply real-identity prelaunch verification. The selected design must complete staged security and entitlement testing under real roles before expanding access, whereas this choice omits that direct protection or verification step.

Question 368

A retiring agent keeps an active service token and searchable sensitive logs after shutdown. Which intervention most directly addresses the failure?

  1. Execute decommissioning that revokes credentials and applies retention/deletion rules.
  2. Use canary rollout, predeclared abort thresholds and tested rollback.
  3. Create change ownership and refresh procedures tied to source updates.
  4. Add operational acceptance gates before calling it production-ready.

Explanation

Correct Answer

A:

Lifecycle governance includes safely removing privileges and artifacts, not merely stopping traffic. Assign operational ownership of secure agent decommissioning and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

B:

The relevant trade-off is not primarily risk-managed deployment lifecycle. A successful fix needs secure agent decommissioning; otherwise secure agent decommissioning remains unverified even after this change.

C:

The control would be justified if continuous knowledge currency were the limiting factor. In this scenario the decisive target is secure agent decommissioning; success requires an architecture that will execute decommissioning that revokes credentials and applies retention/deletion rules.

D:

The option belongs to the broader work on production readiness beyond proof of concept, but it cannot by itself deliver secure agent decommissioning. A decision on this case needs the team to execute decommissioning that revokes credentials and applies retention/deletion rules and verify that outcome.

Question 369

A stakeholder changes target use cases mid-build but no one updates architecture acceptance criteria. What should the solution architect prioritize?

  1. Triage incidents into versioned regression cases with prioritized remediation.
  2. Add operational acceptance gates before calling it production-ready.
  3. Rebaseline scope, measurable outcomes and affected controls through formal change decisions.
  4. Create change ownership and refresh procedures tied to source updates.

Explanation

Correct Answer

C:

Requirement changes can invalidate design and evaluation assumptions even when code still works. Assign operational ownership of controlled scope evolution and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

The proposed action prioritizes incident-driven evaluation improvement. That can coexist with the right solution, but it cannot replace controlled scope evolution: the application must still rebaseline scope, measurable outcomes and affected controls through formal change decisions.

B:

The design concern is real but different: production readiness beyond proof of concept. It does not impose the specific requirement for controlled scope evolution; a suitable solution must rebaseline scope, measurable outcomes and affected controls through formal change decisions and demonstrate the result.

D:

That choice would be responsive to continuous knowledge currency, but the reported defect is different. It leaves controlled scope evolution unproved because there is no commitment to rebaseline scope, measurable outcomes and affected controls through formal change decisions.

Question 370

A customer complains about AI-generated information but support has no route to the original evidence and decision trace. Which design choice best matches the stated constraint?

  1. Triage incidents into versioned regression cases with prioritized remediation.
  2. Add operational acceptance gates before calling it production-ready.
  3. Use prioritized intake categories with owners, severity and response SLAs.
  4. Implement a governed complaint and trace-reconstruction process.

Explanation

Correct Answer

D:

Operational learning and accountability require linking user reports to the responsible system state. Assign operational ownership of audit-supported customer remediation and exercise fallback or rollback during staged deployment, not only after a serious incident.

Incorrect Answers

A:

Improving incident-driven evaluation improvement does not imply audit-supported customer remediation. The selected design must implement a governed complaint and trace-reconstruction process, whereas this choice omits that direct protection or verification step.

B:

This recommendation targets production readiness beyond proof of concept at a different point in the workflow. It leaves audit-supported customer remediation outside the design contract and does not require the team to implement a governed complaint and trace-reconstruction process.

C:

This is a valid architectural concern when optimizing structured post-launch feedback triage. Here it leaves audit-supported customer remediation unaddressed, whereas the intended change must implement a governed complaint and trace-reconstruction process before the case can be considered resolved.

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