Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 9

Question 80

A support-orchestration service decides how to pass tasks between specialists. A customer intent is classified, but no worker has been chosen. Which change most directly addresses this condition?

  1. Associate delegated jobs with stable identifiers and recoverable progress
  2. Require both verified prerequisite states before dispatching the dependent worker
  3. Select a named specialist based on intent and permission boundaries
  4. Validate required fields before invoking the specialist
  5. Treat the unrecognized status as invalid and request clarification or manual review

Explanation

Correct Answer

C:

An explicit dispatch decision links a validated customer need to a worker with relevant expertise and tool access. For the stated case, the check is whether the system can make task routing explicit; this approach supplies that control.

Incorrect Answers

A:

This choice is relevant when the coordinator wants to know whether a stalled task can be resumed; a durable job identity connects resumes to prior work and avoids launching duplicate specialists blindly. Here, however, a customer intent is classified, but no worker has been chosen.

B:

A conjunction of prerequisites prevents the next stage from treating one successful check as equivalent to full authorization; it protects against the case where a task needs both a billing result and an identity confirmation, rather than the failure described here: a customer intent is classified, but no worker has been chosen.

D:

It would be appropriate when a specialist rejects an input because it lacks a case identifier; input checks at the coordinator boundary reduce wasted agent turns and expose missing fields early. Here, a customer intent is classified, but no worker has been chosen, so a different mechanism is required.

E:

A closed status enumeration prevents unfamiliar values from accidentally being treated as success; the alternative problem occurs when a delegated task completes but its result schema has an unknown status. The present scenario instead requires the system to make task routing explicit.

Question 81

An organization uses Claude to summarize an archive of technical incidents. The workload has many independent records and no user waiting for each response. What is the most defensible next step?

  1. Use batch processing and an output manifest for the incident collection
  2. Include stable incident IDs in each request and result mapping
  3. Verify stored document digest and cited spans before accepting the result
  4. Use a synchronous, reviewable workflow for that specific incident

Explanation

Correct Answer

A:

Large independent jobs can be processed without tying up a conversational request; the manifest preserves record identity.

Incorrect Answers

B:

It would be appropriate when a run returns finished summaries but no mapping to original incident IDs; without correlation keys, a bulk response can be structurally complete but impossible to assign safely to the correct incident. Here, the workload has many independent records and no user waiting for each response, so a different mechanism is required.

C:

Matching output to the right input is a semantic and provenance requirement, not merely a batch-completion check; the alternative problem occurs when one summary appears to describe a different source document. The present scenario instead requires the system to favor a controlled offline job over interactive chaining.

D:

This choice is relevant when a critical incident needs an urgent follow-up question from a human; interactive escalation has different latency and control requirements from an archival bulk job. Here, however, the workload has many independent records and no user waiting for each response.

Question 82

A document-processing agent invokes external services while extracting invoice fields. The OCR service sometimes returns an empty string rather than recognized text. Which approach satisfies the stated constraint?

  1. Track pending work in durable application state until the actual result arrives
  2. Validate tool output against the expected result contract before reasoning
  3. Apply a retry cap and bounded backoff outside the model
  4. Validate tool-call inputs against the declared argument schema

Explanation

Correct Answer

B:

An empty result may reflect OCR failure, not an intentionally blank invoice field; validation stops it being promoted to fact.

Incorrect Answers

A:

A long-running call must not be represented as success before completion, especially where later steps depend on its output; the alternative problem occurs when tool execution takes several minutes, but the model has already emitted the request. The present scenario instead requires the system to distinguish missing evidence from a valid empty field.

C:

An application-owned retry policy prevents infinite loops and makes downstream reliability predictable under outages; it protects against the case where an extraction workflow repeats one expensive tool call after a transient error, rather than the failure described here: the OCR service sometimes returns an empty string rather than recognized text.

D:

This choice is relevant when a tool schema requires a numeric invoice total but Claude supplies a string; schema checking at the host boundary prevents type errors and unsafe side effects from malformed model-generated arguments. Here, however, the OCR service sometimes returns an empty string rather than recognized text.

Question 83

A research MCP server uses structured errors and pagination. A returned cursor has expired between paginated requests. Which mechanism should be selected?

  1. Return a distinct expired-cursor error with a restart strategy
  2. Return successful index results plus per-source error statuses
  3. Expose a sanitized error code and trace ID while logging restricted details internally
  4. Return a timeout category with correlation ID and bounded retry guidance

Explanation

Correct Answer

A:

An invalid cursor is not 'no more results'; preserving the error prevents a partial search from appearing complete.

Incorrect Answers

B:

A mixed outcome should retain usable evidence while informing the coordinator that coverage is incomplete; the alternative problem occurs when only one of five federated indexes fails while others succeed. The present scenario instead requires the system to tell the client how to restart pagination safely.

C:

This choice is relevant when a request fails with an internal stack trace containing server paths; structured diagnostics allow troubleshooting without dumping sensitive service internals into a language-model conversation. Here, however, a returned cursor has expired between paginated requests.

D:

Time limits protect system capacity; reporting an empty search result would conceal that no reliable search completed; it protects against the case where an upstream search endpoint times out after several seconds, rather than the failure described here: a returned cursor has expired between paginated requests.

Question 84

A support assistant assigns risk and confidence to proposed customer actions. A high-confidence model verdict has no linked eligibility evidence. What should be implemented at the appropriate boundary?

  1. Audit a representative sample of routine resolutions while reviewing high-risk exceptions individually
  2. Require actual policy checks and relevant source fields before approving action
  3. Compare confidence buckets against observed correctness and adjust routing thresholds
  4. Request the missing field or hand off instead of guessing

Explanation

Correct Answer

B:

Confidence scores are not evidence; external authorization and validated facts determine whether a sensitive operation may proceed.

Incorrect Answers

A:

It would be appropriate when low-risk policy responses need ongoing quality monitoring, but reviewing every response is overwhelming the support team; risk-stratified sampling gives reviewers evidence of routine decision quality without making every low-risk response wait for approval; mandatory high-risk gates remain in place. Here, a high-confidence model verdict has no linked eligibility evidence, so a different mechanism is required.

C:

This choice is relevant when monitoring shows confident unsupported answers increasing after a prompt change; reliability requires empirical error measurement; a fluent explanation is not proof that reported certainty matches accuracy. Here, however, a high-confidence model verdict has no linked eligibility evidence.

D:

Uncertainty arising from missing information needs a data remedy, not stronger prompt language or fabricated certainty; it protects against the case where a decision has low confidence because required account information is missing, rather than the failure described here: a high-confidence model verdict has no linked eligibility evidence.

Question 85

A software team builds a two-pass architecture for Claude-generated code changes. The same agent that writes a patch also certifies that it is safe. Which option best meets the operational need?

  1. Provide the concrete patch, affected code, and measurable review checklist
  2. Return specific failing criteria and request a bounded patch update
  3. Use independent tests and a distinct reviewer for security-sensitive patches
  4. Compare both claims to the public API contract and escalate unresolved risk
  5. Require maintainer authorization after automated review and testing

Explanation

Correct Answer

C:

Self-certification can reproduce the generator's blind spots; independent evidence-based review reduces correlated error.

Incorrect Answers

A:

A reviewer needs artifact evidence and evaluation standards to assess behavior, not a vague description from the writer; it protects against the case where the reviewer says 'looks good' without checking the actual diff, rather than the failure described here: the same agent that writes a patch also certifies that it is safe.

B:

It would be appropriate when a reviewer flags missing tests but the generator reruns without using the feedback; a corrective loop improves only if the generator receives actionable evidence rather than a generic negative verdict. Here, the same agent that writes a patch also certifies that it is safe, so a different mechanism is required.

D:

Conflicting model opinions should be resolved using independent criteria and executable evidence, not majority fluency; the alternative problem occurs when two reviewers produce conflicting assessments of a breaking API change. The present scenario instead requires the system to separate generation from acceptance.

E:

This choice is relevant when the final release decision must be owned by a named maintainer; automation may gather evidence, but the stated governance rule makes a person accountable for promotion to production. Here, however, the same agent that writes a patch also certifies that it is safe.

Question 86

A code review automation runs Claude Code on inbound pull requests. The pipeline must analyze a PR without altering its contents. What is the most appropriate design choice?

  1. Switch from review to authorized bounded editing and validation
  2. Run read-only inspection and feedback generation without write-capable tools
  3. Use an interview or planning step to collect expected behavior
  4. Run allowed checks and inspect logs before proposing edits

Explanation

Correct Answer

B:

A CI reviewer should not silently modify a contributor's branch while supposedly assessing it. This matches the workflow goal to use a review-only execution posture rather than introducing an unrelated action.

Incorrect Answers

A:

Separation between analysis and implementation gives the maintainer control over when automated modifications occur; it protects against the case where a human explicitly requests a patch after reviewing generated suggestions, rather than the failure described here: the pipeline must analyze a PR without altering its contents.

C:

It would be appropriate when an agent is asked to review a feature without knowing its acceptance criteria; undefined success conditions make a code change hard to test; targeted questions can establish the missing contract. Here, the pipeline must analyze a PR without altering its contents, so a different mechanism is required.

D:

This choice is relevant when a test suite is failing and the team wants diagnosis only; diagnostic evidence is needed to locate the cause; granting automatic write access in the discovery phase increases change risk. Here, however, the pipeline must analyze a PR without altering its contents.

Question 87

A team maintains path-aware architecture conventions in Claude Code. The mobile app follows a different package layout from the server. Which response is most reliable?

  1. Give mobile and server trees distinct scoped guidance
  2. Revise path patterns and test their effective selection
  3. Attach the review procedure to infrastructure configuration patterns
  4. Keep shared guidance in one place and scoped exceptions explicitly narrow

Explanation

Correct Answer

A:

Relevant architecture instructions reduce accidental cross-project assumptions when a monorepo contains multiple build systems.

Incorrect Answers

B:

It would be appropriate when a new package directory does not match the existing rules; rule files are not intelligent about renamed folders; matching must be maintained alongside structural refactors. Here, the mobile app follows a different package layout from the server, so a different mechanism is required.

C:

A path-aware rule gives extra caution where necessary without slowing unrelated text edits; it protects against the case where only infrastructure-as-code files require a plan-before-change policy, rather than the failure described here: the mobile app follows a different package layout from the server.

D:

Redundant copies can drift; exceptions should state exactly what changes for a subset of files; the alternative problem occurs when a scoped rule repeats a contradictory general rule verbatim. The present scenario instead requires the system to associate architecture rules with app directories.

Question 88

A development team wants consistent Claude Code behavior across a shared monorepo. Every contributor must follow the same build and review commands for the repository. What should the team do to resolve this particular gap?

  1. Place concise project instructions in a committed project CLAUDE.md file
  2. Place package-specific guidance near the relevant subproject or use path-scoped rules
  3. Inspect project and user guidance together and resolve conflicting directions explicitly
  4. Keep the personal preference in user-level Claude Code guidance

Explanation

Correct Answer

A:

Project instructions make conventions available when working in the repository; they should describe relevant workflow rules rather than replicate all documentation.

Incorrect Answers

B:

It would be appropriate when a single package has unusual testing requirements not relevant elsewhere; local rules reduce noise in unrelated tasks and clarify when a special test command actually applies. Here, every contributor must follow the same build and review commands for the repository, so a different mechanism is required.

C:

This choice is relevant when the team changes a shared convention but contributors have conflicting personal notes; overlapping memory files may supply inconsistent directions; an architect should make shared requirements clear rather than hope the model infers priority. Here, however, every contributor must follow the same build and review commands for the repository.

D:

User-level configuration supports individual habits without imposing personal formatting conventions on the entire repository; it protects against the case where one developer prefers a personal response style that colleagues do not share, rather than the failure described here: every contributor must follow the same build and review commands for the repository.

Question 89

A multi-agent research product must collect, compare, and publish findings. Search, quantitative analysis, and writing each require different expertise and inputs. Which control is most directly relevant?

  1. Run source-search subtasks concurrently and join their evidence before synthesis
  2. Ask the coordinator to compare provenance and escalate unresolved conflicts
  3. Use a coordinator to delegate scoped jobs to search, analysis, and writing workers
  4. Give the coordinator routing and synthesis responsibilities, not a second full search workload

Explanation

Correct Answer

C:

A central orchestrator can assign responsibilities and synthesize outputs while specialists work with only relevant tools and context.

Incorrect Answers

A:

Independent searches have no ordering dependency; parallel fan-out reduces latency while a later merge can still compare findings; it protects against the case where three independent sources must be examined before a combined report is written, rather than the failure described here: search, quantitative analysis, and writing each require different expertise and inputs.

B:

This choice is relevant when two specialists disagree over a key statistic from different documents; the coordinator should reconcile using source quality and dates, not silently select whichever worker sounds more confident. Here, however, search, quantitative analysis, and writing each require different expertise and inputs.

D:

Overlapping roles waste tokens and introduce conflicting versions of evidence; ownership should be explicit in the orchestration contract; the alternative problem occurs when a coordinator routinely duplicates its specialists' detailed searches. The present scenario instead requires the system to coordinate distinct specialist responsibilities.

Question 90

An agentic research pipeline must propagate failures through several workers. A search specialist returns only two of four required evidence sources. How should the workflow be adjusted?

  1. Mark the calculation blocked and include the validation failure
  2. Refuse publication until mandatory worker outcomes are successful or formally waived
  3. Pass a typed timeout outcome instead of a successful empty payload
  4. Return partial-result status and the identifiers of missing sources
  5. Return per-source status with valid outputs and a partial-completion flag

Explanation

Correct Answer

D:

Downstream synthesis must know that evidence collection is incomplete; treating a partial list as full coverage can distort conclusions.

Incorrect Answers

A:

A computed value obtained from unverified inputs should not be promoted as evidence merely because a number was produced; it protects against the case where an analysis agent reports a numeric result but its input validation failed, rather than the failure described here: a search specialist returns only two of four required evidence sources.

B:

This choice is relevant when the final publication stage needs all required source checks complete; a release gate should inspect explicit prerequisite state rather than trust an optimistic synthesis paragraph. Here, however, a search specialist returns only two of four required evidence sources.

C:

It would be appropriate when a writing agent receives an upstream timeout labeled success; correct error semantics prevent downstream agents from mistaking unavailable data for evidence of absence. Here, a search specialist returns only two of four required evidence sources, so a different mechanism is required.

E:

An all-or-nothing summary discards useful work, while an all-success label conceals missing coverage; partial status preserves both truths; the alternative problem occurs when one of three independent data sources fails while the others are usable. The present scenario instead requires the system to represent incomplete coverage explicitly.

Claude Certified Architect - Foundations (CCAR-F) (CCA-F) Practice Tests