Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 14

Question 130

A company processes a heavy morning batch of supplier certificates, with individual analysis taking much longer than its web request timeout. Which intervention is most appropriate for this use case?

  1. Validate model output against a business schema before side effects
  2. Use a queue and asynchronous workers for bursty workloads
  3. Capture reviewer corrections as labeled feedback
  4. Persist workflow state and make retries idempotent

Explanation

Correct Answer

B:

Decouple intake from processing through durable queues, bounded workers, and status callbacks rather than keeping user connections open; measuring backlog and completion under load tests the exact burst tolerance the design promises; evaluate the design with job completion rate and queue age during the peak intake window.

Incorrect Answers

A:

Use structured output parsing, semantic validation, and a separate authorized command handler so generated text is never treated directly as a write instruction; output parsing constrains generated data but does not by itself manage retry idempotency or transport durability; this alternative would measure invalid write payloads rejected before execution, with zero partial supplier mutations; it does not establish job completion rate and queue age during the peak intake window for the described use case.

C:

Preserve authorized reviewer labels linked to case versions and turn them into regression examples with data-retention controls; a feedback dataset improves future evaluation; it does not guarantee that the current response is timely or properly permissioned; this alternative would measure regression pass rate on previously corrected clause-omission cases; it does not establish job completion rate and queue age during the peak intake window for the described use case.

D:

Persist state transitions and use idempotency keys at side-effect boundaries to avoid duplicate transactions across replays; durable-state control prevents duplicate effects but cannot decide whether a refund is substantively authorized; this alternative would measure repeated and interrupted test runs yielding exactly one committed credit per approved case; it does not establish job completion rate and queue age during the peak intake window for the described use case.

Question 131

A short-lived cache frequently expires before the next eligible request and write cost dominates expected savings. What is the most direct remediation?

  1. Instrument cache effectiveness with token usage counters
  2. Modularize shared task instructions for consistent reuse
  3. Compare cache lifetime with actual reuse patterns
  4. Keep dynamic fields out of cached prefix content

Explanation

Correct Answer

C:

Assess eligible caching windows, traffic frequency, and write/read economics against real reuse intervals before selecting a cache policy; reuse-adjusted economics tests whether the caching decision pays off under actual traffic; verify the repair using net cost per accepted request across realistic interarrival times.

Incorrect Answers

A:

A faster model might improve latency but would not explain a missing cache hit on identical prompts; compare cache-creation and cache-read token telemetry and inspect prefix ordering, eligible size, and reuse timing; that control is assessed by sustained cache-read token share under repeated workload replay; the incident instead requires net cost per accepted request across realistic interarrival times.

B:

Prompt modularity simplifies governance; it cannot alone guarantee that a sensitive tool call is authorized; maintain a versioned reusable prompt module with explicit parameters and regression tests across consumers; that control is assessed by cross-application prompt-version alignment and regression pass rate; the incident instead requires net cost per accepted request across realistic interarrival times.

D:

A cache breakpoint cannot compensate for changes to earlier content in the hashed prefix; move changing metadata after the stable prefix so cache keys remain reusable across otherwise similar requests; that control is assessed by prefix reuse ratio when request metadata changes across calls; the incident instead requires net cost per accepted request across realistic interarrival times.

Question 132

A vendor research assistant retrieves hostile third-party web pages during an authorized sourcing task. Security tests attempt to make those pages trigger outbound disclosure. Which measured result proves the boundary holds?

  1. Zero payment writes when tested with the analytics identity and forged instructions
  2. Denial of unauthorized actions and harmful outputs across multi-stage red-team paths
  3. Attack success rate on adversarial web pages and zero unauthorized egress calls
  4. Forbidden file and network access attempts blocked in execution tests

Explanation

Correct Answer

C:

Simulated hostile documents test the trust boundary directly; measure attack success rate on adversarial web pages and zero unauthorized egress calls at the release gate; keep source contents in a lower-trust evidence role and prevent embedded commands from crossing into tool authority.

Incorrect Answers

A:

Direct permission tests verify security independent of model obedience; enforce role and scope in the backend execution service using credentials that cannot perform the forbidden action; this tests zero payment writes when tested with the analytics identity and forged instructions; the actual release condition is attack success rate on adversarial web pages and zero unauthorized egress calls.

B:

End-to-end adversarial cases assess the combined defenses and their failure isolation; layer input handling, model instructions, backend authorization, schema checks, and output review so one failed defense does not grant action authority; this tests denial of unauthorized actions and harmful outputs across multi-stage red-team paths; the actual release condition is attack success rate on adversarial web pages and zero unauthorized egress calls.

D:

Sandbox escapes and denial logs are more relevant than the assistant's stated intentions; isolate execution with restricted filesystem, network, credentials, resources, and observable policy denials; this tests forbidden file and network access attempts blocked in execution tests; the actual release condition is attack success rate on adversarial web pages and zero unauthorized egress calls.

Question 133

A certificate-processing service is being moved behind an asynchronous queue so that peak-hour browser requests do not hold long-running jobs open. Which operational measurement should determine whether the redesign works?

  1. Regression pass rate on previously corrected clause-omission cases
  2. Job completion rate and queue age during the peak intake window
  3. Invalid write payloads rejected before execution, with zero partial supplier mutations
  4. Percentage of ingested documents meeting schema checks before any Claude call

Explanation

Correct Answer

B:

Measuring backlog and completion under load tests the exact burst tolerance the design promises; measure job completion rate and queue age during the peak intake window at the release gate; decouple intake from processing through durable queues, bounded workers, and status callbacks rather than keeping user connections open.

Incorrect Answers

A:

Retesting concrete corrections detects whether a release reintroduces known business mistakes; preserve authorized reviewer labels linked to case versions and turn them into regression examples with data-retention controls; this tests regression pass rate on previously corrected clause-omission cases; the actual release condition is job completion rate and queue age during the peak intake window.

C:

Pre-write enforcement proves that the model cannot bypass the target system's integrity contract; use structured output parsing, semantic validation, and a separate authorized command handler so generated text is never treated directly as a write instruction; this tests invalid write payloads rejected before execution, with zero partial supplier mutations; the actual release condition is job completion rate and queue age during the peak intake window.

D:

A schema-conformance gate reveals extraction failures before they can contaminate reasoning; an input-adapter layer validates document type, field types, provenance, and required metadata before the model reasons over values; this tests percentage of ingested documents meeting schema checks before any Claude call; the actual release condition is job completion rate and queue age during the peak intake window.

Question 134

The model quotes a retired edition because a chunk lacks its effective date and document identifier. What is the most direct remediation?

  1. Allocate input and output token budgets per workflow step
  2. Compact extended sessions without discarding commitments
  3. Persist operational decisions outside transient chat history
  4. Attach provenance metadata to selected context

Explanation

Correct Answer

D:

Supply source identity, version, effective date, and retrieval relevance alongside each chunk so the model can attribute evidence correctly; metadata-aware checks distinguish legitimate historic references from current-policy answers; verify the repair using citation correctness by effective-date version in temporal regression questions.

Incorrect Answers

A:

Merely expanding every context window can increase cost while leaving output truncation unaddressed; reserve output capacity and bound each stage's retrieved context and response length according to the deliverable structure; that control is assessed by complete mandatory report sections under the worst supported input size; the incident instead requires citation correctness by effective-date version in temporal regression questions.

B:

Blind summarization can lose critical exceptions unless durable commitments are explicitly represented; store a structured, reviewable summary of decisions, open issues, and provenance; inject it at checkpoints rather than replaying every chat turn; that control is assessed by retention of signed-off requirements after long simulated conversations; the incident instead requires citation correctness by effective-date version in temporal regression questions.

C:

Long context alone cannot provide transactional durability or a trustworthy authorization trail; store approvals and workflow state in an authoritative durable system; treat chat history as context, not a transaction ledger; that control is assessed by correct recovery of approvals and pending actions after forced session reset; the incident instead requires citation correctness by effective-date version in temporal regression questions.

Question 135

A service processes easy informational queries and rare, consequential disputes. The team proposes separate model routes. Which comparison would prove that routing preserves risk-tier quality while reducing total spend?

  1. Successful completion of critical workflows during injected model outages
  2. End-to-end completed cases per paid work-hour with rework and appeal volumes
  3. End-to-end p95 latency and error rate against the agreed service objective
  4. Quality and total cost segmented by request risk tier under a fixed accuracy floor

Explanation

Correct Answer

D:

Segmented evaluation reveals whether additional reasoning spend is targeted at cases that benefit; measure quality and total cost segmented by request risk tier under a fixed accuracy floor at the release gate; route by validated task complexity and consequence, using stronger reasoning or human review where evidence justifies the incremental spend.

Incorrect Answers

A:

Fallback drills demonstrate that the business process is not hostage to one generative dependency; separate critical deterministic operations from AI enhancement and provide explicit fallback and manual paths when model service is unavailable; this tests successful completion of critical workflows during injected model outages; the actual release condition is quality and total cost segmented by request risk tier under a fixed accuracy floor.

B:

Net throughput reflects business output rather than the volume of AI text produced; measure net time saved after review, appeals, and escalations, segmented by complexity, before claiming productivity; this tests end-to-end completed cases per paid work-hour with rework and appeal volumes; the actual release condition is quality and total cost segmented by request risk tier under a fixed accuracy floor.

C:

Tail measurement detects the bad experiences that averages can obscure; define a p95 or p99 target against real end-to-end requests, including retrieval and tools, rather than reporting an average model time; this tests end-to-end p95 latency and error rate against the agreed service objective; the actual release condition is quality and total cost segmented by request risk tier under a fixed accuracy floor.

Question 136

A pilot is transferred from architects to an internal support team unfamiliar with escalation and retry behavior. Which approach best meets the stated requirement?

  1. Publish an incident runbook with containment and rollback steps
  2. Version prompts, model routes, tools, and evaluation artifacts together
  3. Specify tool schemas, errors, and authorization contracts
  4. Maintain an end-to-end data-flow and trust-boundary diagram
  5. Prepare operator onboarding and support handoff materials

Explanation

Correct Answer

E:

Provide system overview, access setup, troubleshooting flowcharts, monitoring thresholds, and named support responsibilities; independent operator readiness is stronger evidence than architect-led demonstrations; evaluate the design with successful completion of standard support drills by staff who did not build the system.

Incorrect Answers

A:

Maintain runbooks for detection, containment, safe fallback, rollback, and communication with named incident roles; a high-level diagram alone is not an executable incident response procedure; this alternative would measure successful containment time in a simulated stale-index incident; it does not establish successful completion of standard support drills by staff who did not build the system for the described use case.

B:

Maintain a release manifest linking model configuration, prompt, tool contracts, data versions, and evaluation results; recording only the model name omits other behavior-changing dependencies; this alternative would measure reproducibility of a sampled historical answer using its full manifest; it does not establish successful completion of standard support drills by staff who did not build the system for the described use case.

C:

Define typed inputs, outputs, error semantics, idempotency, required scopes, and audit obligations for tool boundaries; aPI examples without failure and permission semantics leave dangerous integration ambiguity; this alternative would measure consumer contract pass rate under timeouts, invalid arguments, and permission denials; it does not establish successful completion of standard support drills by staff who did not build the system for the described use case.

D:

A current diagram should show data classes, flows, processing boundaries, identities, trust zones, and side effects; a feature list does not reveal crossing of sensitive trust boundaries; this alternative would measure completeness of sensitive-data flow paths identified during security review; it does not establish successful completion of standard support drills by staff who did not build the system for the described use case.

Question 137

In a multi-agent market analysis, specialist drafts are updated concurrently and readers sometimes merge out-of-date versions. A separate investigation manager keeps delegating unanswered questions in a circular chain. Which TWO controls prevent stale handoffs and uncontrolled delegation? Choose TWO.

  1. Version the handoff artifacts between agents
  2. Separate independent critique from generation
  3. Bound handoffs, retries, and recursion across agents
  4. Run independent specialists in parallel with a merge contract
  5. Delegate only specialized subtasks that add measurable value

Explanation

Correct Answers

A:

Immutable, versioned handoff artifacts and typed schemas prevent one agent from consuming ambiguous or mismatched intermediate state; version-aware handoffs can be tested deterministically under racing writes; its corresponding workstream should be validated by zero stale-artifact merges across concurrent revision and retry scenarios.

C:

Enforce maximum depth, handoff count, timeouts, and explicit unresolved-case escalation in the orchestrator; stressing unresolved cases proves that the system can fail predictably and within budget; its corresponding workstream should be validated by zero unbounded cycles under adversarial unresolved tasks while retaining a useful escalation record.

Incorrect Answers

B:

A distinct verifier with a rubric and evidence access can challenge unsupported assertions before an external document is released; a critic agent does not repair missing source documents or guarantee that all facts are verifiable; its natural acceptance measure is verified claim support rate after independent critique on seeded unsupported drafts, which does not satisfy the separate workstreams being assessed.

D:

Parallel specialists can shorten independent investigations, but the product-planning workers here share changing drafts and need versioned handoffs; the second workstream instead needs a cap on recursive delegation, not more parallel branches.

E:

Introduce a specialist only when its independent expertise, tool scope, or parallelism beats a simpler call on representative evaluations; more agents by themselves do not provide additional knowledge, authority, or better outcomes; its natural acceptance measure is marginal quality improvement per added specialist against its incremental cost and latency, which does not satisfy the separate workstreams being assessed.

Question 138

An audit question typically concerns a single control, yet earlier runs included hundreds of irrelevant report pages. Which metric would show that focused retrieval improves useful work per token?

  1. Cost and successful throughput per nightly processing window
  2. Cache-read rate, p95 latency, and cost per accepted answer under repeated queries
  3. Accepted resolution rate and cost per case across routed difficulty levels
  4. Supported-answer rate per input token on labeled audit questions

Explanation

Correct Answer

D:

Evidence efficiency tests whether trimming preserves the information that matters; measure supported-answer rate per input token on labeled audit questions at the release gate; retrieve relevant excerpts and retain enough surrounding evidence for interpretation rather than indiscriminately filling the context window.

Incorrect Answers

A:

Deadline-oriented throughput is the relevant performance objective for noninteractive work; use suitable batch orchestration and queue control to trade immediate response time for predictable throughput and expense; this tests cost and successful throughput per nightly processing window; the actual release condition is supported-answer rate per input token on labeled audit questions.

B:

Joint economics and quality checks demonstrate that reuse helps without sacrificing correctness; keep a stable eligible prompt prefix and measure cache-read tokens to reduce repeated processing; this tests cache-read rate, p95 latency, and cost per accepted answer under repeated queries; the actual release condition is supported-answer rate per input token on labeled audit questions.

C:

Stratified monitoring confirms economic gains do not displace expensive errors onto users; evaluate a smaller model on straightforward request strata and escalate only where measurable capability is needed; this tests accepted resolution rate and cost per case across routed difficulty levels; the actual release condition is supported-answer rate per input token on labeled audit questions.

Question 139

A low-risk product-description assistant generates thousands of suggestions daily. The risk office is evaluating sample-based review instead of approving every line. Which evidence would establish an acceptable error-detection and workload trade-off?

  1. Recurrence rate of categorized, previously corrected error types
  2. Critical-action dual-approval compliance and zero bypasses during staged tests
  3. Defect-detection rate and reviewer workload under a risk-tiered sampling policy
  4. Correct escalation of ambiguous legal cases with reviewer attribution
  5. Zero disbursements without an auditable transaction-specific approver

Explanation

Correct Answer

C:

Sampling effectiveness must balance residual errors and human capacity; measure defect-detection rate and reviewer workload under a risk-tiered sampling policy at the release gate; combine automated checks with representative human sampling and exception-based review scaled to risk.

Incorrect Answers

A:

Structured override trends reveal whether feedback improves the system over time; record correction reasons, provenance, and case labels under retention rules to support evaluation and process redesign; this tests recurrence rate of categorized, previously corrected error types; the actual release condition is defect-detection rate and reviewer workload under a risk-tiered sampling policy.

B:

Approval testing must target the cases where one mistake affects many users; apply risk-tiered approvals, including dual control or independent sign-off where policy demands, to exceptionally consequential mutations; this tests critical-action dual-approval compliance and zero bypasses during staged tests; the actual release condition is defect-detection rate and reviewer workload under a risk-tiered sampling policy.

D:

A consequence-aware test checks whether unsupported authoritative decisions are avoided; define escalation thresholds and route consequential interpretations to qualified legal or compliance owners with source context; this tests correct escalation of ambiguous legal cases with reviewer attribution; the actual release condition is defect-detection rate and reviewer workload under a risk-tiered sampling policy.

E:

Payment simulation tests prove approvals are bound to the exact side effect; separate recommendation from execution and require an authenticated human approval bound to the exact amount, payee, and transaction; this tests zero disbursements without an auditable transaction-specific approver; the actual release condition is defect-detection rate and reviewer workload under a risk-tiered sampling policy.

Question 140

A sales assistant serves multiple account managers whose CRM permissions differ by region. What is the most defensible next step?

  1. Enforce authorization inside the tool backend
  2. Require separate approval for high-impact mutations
  3. Verify tenant ownership of resource identifiers
  4. Use delegated user identity for per-user data access
  5. Record the user, tenant, and action in audit trails
  6. Issue short-lived scoped credentials for tools

Explanation

Correct Answer

D:

Carry an appropriate user-bound authorization context to downstream services and enforce record-level access per request; user-context tests verify enforcement of the source system's individual access rules; evaluate the design with zero cross-user CRM record disclosures in permission-matrix tests.

Incorrect Answers

A:

Implement enforceable role, amount, and resource checks at the action service, not just in natural-language instructions; prompt-only restrictions can be bypassed and are not an independently audited authorization control; this alternative would measure rejection of unauthorized tool calls even when sent outside the model workflow; it does not establish zero cross-user CRM record disclosures in permission-matrix tests for the described use case.

B:

Bind high-impact actions to a separately verified approval and narrowly scoped execution credential; a general sign-in session does not imply user consent to every destructive operation in that session; this alternative would measure zero critical network changes without a recorded action-specific approval; it does not establish zero cross-user CRM record disclosures in permission-matrix tests for the described use case.

C:

Check the requested resource against the authenticated tenant and role at the data service rather than trusting user-provided identifiers; syntactically valid IDs and a logged-in session do not establish ownership of a resource; this alternative would measure denial of cross-tenant identifier enumeration attempts in direct tool tests; it does not establish zero cross-user CRM record disclosures in permission-matrix tests for the described use case.

E:

Propagate authenticated actor, tenant, delegated scope, request ID, and action outcome into tamper-resistant audit records; logging only a model response cannot attribute which authenticated principal caused a backend change; this alternative would measure audit reconstruction of actor, authority, target, and result for sampled mutations; it does not establish zero cross-user CRM record disclosures in permission-matrix tests for the described use case.

F:

Prefer tightly scoped, expiring credentials with secure rotation and audited issuance for the precise operation; encrypting a shared permanent key at rest does not mitigate broad privilege after the key is exposed; this alternative would measure credential-expiry and scope-denial tests after simulated token compromise; it does not establish zero cross-user CRM record disclosures in permission-matrix tests for the described use case.

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