Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 32

Question 310

Claude Code is used to modernize a legacy service under production constraints. The team disagrees about the safest migration approach. Which control is most directly relevant?

  1. Generate alternative plans with trade-offs and identify missing evidence
  2. Inspect dependencies and outline migration, rollback, and validation steps in plan mode
  3. Limit writes to authorized paths and escalate the out-of-scope changes
  4. Search for the flag and inspect references using read-only tools
  5. Edit the identified key and run the acceptance test

Explanation

Correct Answer

A:

A decision stage should expose risks and open questions rather than let the assistant silently pick an unreviewed migration strategy.

Incorrect Answers

B:

For the separate situation where the requested change affects database schema and multiple integrations, high-impact changes benefit from a reviewable plan that exposes assumptions and ordering before writes occur; the current task is to compare proposals before committing to one.

C:

Approval for a plan does not grant unrestricted filesystem rights; actual tool permissions must still match the authorized work; the alternative would make sense if the final plan is accepted, but some files lie outside the permitted workspace. It does not directly fix the observed condition: the team disagrees about the safest migration approach.

D:

This control matters when the task is to locate files referencing an old environment flag; locating references is a discovery task; executable patches are premature until intended behavior is understood. The required decision is different: the team disagrees about the safest migration approach.

E:

A concrete low-risk request with clear verification can move directly to execution without an elaborate architecture phase; that assumes a one-line configuration adjustment has an approved target and acceptance check. The item tests the different ability to compare proposals before committing to one.

Question 311

A research agent must explore conflicting hypotheses with limited resources. The coordinator keeps asking new workers for the same question. How should the workflow be adjusted?

  1. Allow a bounded agent loop to generate follow-ups from validated findings
  2. Assign explicit confirmation and challenge tasks before the final synthesis
  3. Track completed task intents and reuse valid existing findings
  4. Enumerate documents and allocate bounded extraction tasks to each

Explanation

Correct Answer

C:

Deduplicating work prevents an apparent multi-agent design from simply repeating the same investigation. The relevant requirement is to detect redundant delegated subtasks, and this measure operates at that decision point.

Incorrect Answers

A:

This control matters when an initial article may reveal unknown organizations that require follow-up queries; uncertain discovery benefits from adaptive next steps; the budget and tools still require host control. The required decision is different: the coordinator keeps asking new workers for the same question.

B:

Separate search objectives reduce confirmation bias and give the coordinator grounds to compare conflicting material; that assumes an answer requires independently checking both supportive and contradictory evidence. The item tests the different ability to detect redundant delegated subtasks.

D:

For the separate situation where the search space has a fixed list of ten known documents, a finite inventory can be processed predictably without an autonomous planner deciding what to look at next; the current task is to detect redundant delegated subtasks.

Question 312

A team maintains path-aware architecture conventions in Claude Code. A scoped rule repeats a contradictory general rule verbatim. Which implementation decision is the best fit?

  1. Revise path patterns and test their effective selection
  2. Give mobile and server trees distinct scoped guidance
  3. Constrain edit permissions to approved paths at the execution layer
  4. Keep shared guidance in one place and scoped exceptions explicitly narrow

Explanation

Correct Answer

D:

Redundant copies can drift; exceptions should state exactly what changes for a subset of files. This matches the workflow goal to remove redundant conflicting copies rather than introducing an unrelated action.

Incorrect Answers

A:

Rule files are not intelligent about renamed folders; matching must be maintained alongside structural refactors; that assumes a new package directory does not match the existing rules. The item tests the different ability to remove redundant conflicting copies.

B:

For the separate situation where the mobile app follows a different package layout from the server, relevant architecture instructions reduce accidental cross-project assumptions when a monorepo contains multiple build systems; the current task is to remove redundant conflicting copies.

C:

Scoped instruction text improves intent but does not prevent a capable tool from writing elsewhere when misused; the alternative would make sense if a critical write restriction is written only as path-specific prose. It does not directly fix the observed condition: a scoped rule repeats a contradictory general rule verbatim.

Question 313

A document-processing architecture decides between online calls and batch jobs. A large job contains some documents that require human review. What should the architect specify in the design?

  1. Invoke the model synchronously with bounded timeout and fallback
  2. Maintain stable item IDs and a durable submission ledger
  3. Mark uncertain items for separate human adjudication while preserving successful outputs
  4. Use batch requests with durable tracking and completion monitoring

Explanation

Correct Answer

C:

Partial automation can still be useful, but the pipeline must not pretend every item met the same confidence threshold.

Incorrect Answers

A:

This control matters when a cashier needs a tax calculation before finishing the same transaction; a transaction depends on the output now; asynchronous batch scheduling would not satisfy the immediate completion dependency. The required decision is different: a large job contains some documents that require human review.

B:

A replay-safe manifest distinguishes outstanding items from completed or already submitted work and prevents accidental duplicate processing; that assumes a batch worker may retry failed items after network interruption. The item tests the different ability to split automated completion from review status.

D:

For the separate situation where a nightly backlog consists of independent invoices due by the next morning, the deadline tolerates deferred execution and requests are independent, making batch orchestration a suitable operational pattern; the current task is to split automated completion from review status.

Question 314

An MCP service encounters errors while fetching customer records. A tool call is rejected because the caller lacks access to the account. Which change most directly addresses this condition?

  1. Return a distinct not-found status without recommending blind retries
  2. Classify the issue as transient unavailable and provide a safe diagnostic identifier
  3. Return a permission-denied outcome and require an authorized path
  4. Return a structured rate-limit error and retry-after guidance

Explanation

Correct Answer

C:

A permission error cannot be fixed by repeating the same call; it needs identity correction or escalation without leaking account details.

Incorrect Answers

A:

This control matters when a requested account number is well formed but does not exist; a genuinely missing resource is different from transient infrastructure failure; retrying unchanged identifiers adds load without resolving it. The required decision is different: a tool call is rejected because the caller lacks access to the account.

B:

Clear error categorization lets orchestration retry selectively while keeping internal stack traces out of model-facing content; that assumes a database loses connectivity during a lookup. The item tests the different ability to preserve authorization failure semantics.

D:

For the separate situation where a backend returns HTTP 429 with a retry window, throttling is transient and should be handled with bounded delay, not treated as an absent customer record; the current task is to preserve authorization failure semantics.

Question 315

A product team decomposes complex Claude tasks into subtasks. The number of investigative steps cannot be known until initial evidence arrives. What is the most defensible next step?

  1. Extract, validate, calculate, then explain in that order
  2. Collect sources concurrently and synthesize once all results pass validation
  3. Use an agent loop with step budgets and decisions conditioned on observations
  4. Implement a deterministic pipeline around Claude calls rather than a free-form planner
  5. Call Claude once with a precise classification instruction and output contract

Explanation

Correct Answer

C:

Adaptive tool choice is useful for uncertain exploration, provided the host constrains possible actions and termination.

Incorrect Answers

A:

The arithmetic is only meaningful after source values pass type and completeness checks; dependencies determine stage order; that assumes an analysis must validate extracted figures before calculating a ratio. The item tests the different ability to allow controlled adaptive planning.

B:

This control matters when a report must combine three unrelated data sources before drafting; independent lookups can run together, but writing should wait for the join to prevent omissions. The required decision is different: the number of investigative steps cannot be known until initial evidence arrives.

D:

Known stable stages are easier to test and audit as code; open-ended planning adds variance without meeting an unmet requirement; the alternative would make sense if a workflow repeats the same fixed five stages for every case. It does not directly fix the observed condition: the number of investigative steps cannot be known until initial evidence arrives.

E:

For the separate situation where a customer email requires one classification label and no external data, a single well-scoped task avoids unnecessary agent coordination when no independent dependencies or tools are needed; the current task is to allow controlled adaptive planning.

Question 316

An agent controller must coordinate recoverable workflows across interruptions. A new agent task accidentally inherits old customer contact details. Which approach satisfies the stated constraint?

  1. Persist origin session, fork lineage, task ID, and artifact references
  2. Replay the persisted, ID-correlated tool_result into the restored conversation
  3. Create a clean conversation and explicitly scoped context for the new task
  4. Reauthorize actions against the current permission configuration

Explanation

Correct Answer

C:

Starting fresh avoids accidentally promoting data from an unrelated case into a new workflow. For the stated case, the check is whether the system can separate sessions by privacy boundary; this approach supplies that control.

Incorrect Answers

A:

Provenance across resumes and forks is essential for diagnosing which state or branch produced a disputed conclusion; the alternative would make sense if engineers need to trace exactly where a resumed result originated. It does not directly fix the observed condition: a new agent task accidentally inherits old customer contact details.

B:

For the separate situation where a tool result was durably saved with its request ID, but the host crashed before passing it back to Claude, when the result is already durably recorded, the host can resume the exact tool-use/result exchange without repeating the side-effecting operation; the current task is to separate sessions by privacy boundary.

D:

Session continuity is not continuing authorization; permissions may change independently of the saved transcript; that assumes a worker changed permissions in an earlier session, but today's role policy differs. The item tests the different ability to separate sessions by privacy boundary.

Question 317

An engineering assistant uses tool wrappers to standardize behavior. A hook modifies returned data, and support needs to debug discrepancies. Which mechanism should be selected?

  1. Deny mutation tools while the workflow is in the read-only discovery phase
  2. Record the original result reference, transformation version, and normalized output
  3. Normalize both responses to a common file-and-line reference schema
  4. Check expected test coverage or completion markers in addition to the exit code

Explanation

Correct Answer

B:

Traceable normalization allows engineers to determine whether a mismatch arose in the tool, wrapper, or later model reasoning.

Incorrect Answers

A:

Phase-based permissions ensure exploration cannot silently become execution before an explicit workflow transition; the alternative would make sense if a request policy allows read operations but forbids writes during discovery. It does not directly fix the observed condition: a hook modifies returned data, and support needs to debug discrepancies.

C:

For the separate situation where two code-search tools return incompatible source-location formats, common locations make evidence comparable and simplify provenance preservation across different retrieval implementations; the current task is to make transformations observable.

D:

This control matters when a test runner returns a success exit code but logs an incomplete test suite; a zero exit status alone may not prove required checks ran; outcome validation should match the acceptance criteria. The required decision is different: a hook modifies returned data, and support needs to debug discrepancies.

Question 318

An AI research assistant prepares high-stakes organizational recommendations. A minor stylistic issue does not affect factual or safety requirements. What should be implemented at the appropriate boundary?

  1. Request the applicable policy or route to its owner
  2. Present competing interpretations and escalate the trade-off to the responsible decision maker
  3. Apply documented style rules automatically and continue
  4. Mark the affected conclusion as unverified and request further review

Explanation

Correct Answer

C:

Human review capacity is best reserved for consequential ambiguity; deterministic cosmetic corrections can be handled within the workflow.

Incorrect Answers

A:

An agent cannot safely invent an organization's normative requirements from generic patterns; that assumes a requested final verdict depends on a policy that has not been supplied. The item tests the different ability to avoid escalating low-risk formatting decisions.

B:

This control matters when the cited facts support two materially different recommendations; when evidence does not uniquely determine action, the architect should avoid manufacturing unwarranted certainty. The required decision is different: a minor stylistic issue does not affect factual or safety requirements.

D:

For the separate situation where a key source is inaccessible and no equivalent evidence is available, the absence of a crucial source limits what the assistant can responsibly conclude; a plausible summary cannot replace verification; the current task is to avoid escalating low-risk formatting decisions.

Question 319

A coordinator divides a codebase-migration assessment among specialist agents. The coordinator can pass a 200-page repository dump to every worker. Which option best meets the operational need?

  1. Return a structured blocker status with severity and supporting evidence
  2. Define distinct tasks and structured outputs for each worker
  3. Send only relevant file maps and task-specific excerpts to each specialist
  4. Wait for both prerequisite worker results before launching the integration-check task

Explanation

Correct Answer

C:

Broadcasting the whole repository inflates context and cost and obscures what each worker must evaluate. The relevant requirement is to limit redundant context transfer, and this measure operates at that decision point.

Incorrect Answers

A:

A typed blocker can halt or reroute downstream steps; a buried sentence is easy for a later summarizer to miss; that assumes a security worker discovers an immediate blocker during parallel migration analysis. The item tests the different ability to limit redundant context transfer.

B:

For the separate situation where one worker is responsible for dependency scanning and another for test-risk analysis, explicit outputs make the coordinator's merge traceable and prevent the workers from competing over the same incomplete artifact; the current task is to limit redundant context transfer.

D:

A join condition prevents testing a proposed change against an incomplete dependency or migration plan; the alternative would make sense if an integration test needs both migration and dependency analyses to finish. It does not directly fix the observed condition: the coordinator can pass a 200-page repository dump to every worker.

Question 320

A development agent explores competing fixes across multiple sessions. A resumed worker reports success based on a cached test from last week. What is the most appropriate design choice?

  1. Fork a common session context for each experiment
  2. Rerun relevant tests on the current revision before accepting success
  3. Resume the relevant session and inspect the recorded pending work
  4. Preserve goals, constraints, decisions, open issues, and source references while discarding noise
  5. Use separate worktrees or sandboxes for mutable branch experiments

Explanation

Correct Answer

B:

Tests describe a specific code state; a prior green result cannot establish correctness for a changed repository. The relevant requirement is to invalidate stale verification evidence, and this measure operates at that decision point.

Incorrect Answers

A:

This control matters when two proposed fixes need separate experiments from the same baseline; branching keeps baseline assumptions comparable, whereas serially editing one conversation blends conclusions. The required decision is different: a resumed worker reports success based on a cached test from last week.

C:

For the separate situation where a developer needs to continue yesterday's unfinished refactor, continuing the actual conversation preserves intent and decisions, although the current code must still be validated; the current task is to invalidate stale verification evidence.

D:

Structured compaction retains what is needed for continuation and reduces irrelevant history without implying forgotten facts are still available; the alternative would make sense if a session has grown so large that prior assumptions hide the active requirement. It does not directly fix the observed condition: a resumed worker reports success based on a cached test from last week.

E:

Forked dialogue history alone does not isolate disk writes; workspace separation prevents cross-experiment contamination; that assumes one experiment altered local files that another experiment must not see. The item tests the different ability to invalidate stale verification evidence.

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