Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 34

Question 330

A returning support case needs its approved customer preferences but not the entire historical ticket archive. Which of these alternatives serves the stated goal?

  1. Avoid returning unnecessary raw records
  2. Load task-specific memory at the start
  3. Start a fresh isolated context for a subtask
  4. Write durable memory only through an authorized tool

Explanation

Correct Answer

B:

Relevant stored facts are retrieved and inserted into the allowed context for the current task; the context selection must be scoped to the authenticated customer.

Incorrect Answers

A:

The tool should return task-relevant, minimized structured data instead of dumping unneeded sensitive details; by comparison, the intended solution relies on this: relevant stored facts are retrieved and inserted into the allowed context for the current task; excessive outputs increase context costs and exposure while burying the useful evidence.

C:

A separate worker receives only the necessary files and reports a minimal result, whereas the requirement calls for another behavior: relevant stored facts are retrieved and inserted into the allowed context for the current task; isolation only works when tool rights and returned text follow the same boundary.

D:

A storage tool persists a validated fact in an application-owned record; however, the critical mechanism here is that relevant stored facts are retrieved and inserted into the allowed context for the current task, and model response text alone does not update a durable external database.

Question 331

A deployment accidentally omits the policy that restricts destructive tool execution. Which design would meet this acceptance criterion?

  1. Keep secrets out of settings files committed to Git
  2. Validate configuration at startup
  3. Fail closed on missing safety configuration
  4. Check effective settings across scopes

Explanation

Correct Answer

C:

Startup validation blocks unsafe execution rather than falling back to all permissions—with this limitation: the absence of a guardrail must not silently widen access.

Incorrect Answers

A:

Secrets belong in protected secret stores or environment injection rather than tracked source, whereas the requirement calls for another behavior: startup validation blocks unsafe execution rather than falling back to all permissions; repository history can preserve credentials even after a later edit deletes them.

B:

Early checks fail fast with a clear operator error before the service accepts work; however, the critical mechanism here is that startup validation blocks unsafe execution rather than falling back to all permissions, and deferring discovery to live traffic increases incident impact.

D:

Resolving effective configuration reveals higher-priority policy and machine-specific overrides; this cannot substitute for the behavior that startup validation blocks unsafe execution rather than falling back to all permissions, because reading only one settings file can miss the operative restriction.

Question 332

A calendar service returns a permission denial and the assistant must not assume a meeting was booked. Which technical control is appropriate?

  1. Report tool failures explicitly
  2. Validate and execute the tool outside the model
  3. Preserve tool error codes separately from text
  4. Return a matching tool_result block

Explanation

Correct Answer

A:

A structured error result communicates the failed operation and supports safe recovery; notably, silently replacing an error with an empty success encourages false completion claims.

Incorrect Answers

B:

Application code enforces permission checks and performs the actual side effect or read; in contrast, a structured error result communicates the failed operation and supports safe recovery, while the presence of tool_use content is not evidence the backend was queried.

C:

Typed errors guide safe retries, clarification, or escalation without inventing data; this addresses a separate concern, not the fact that a structured error result communicates the failed operation and supports safe recovery; free-form error prose can be misread or stripped by downstream clients.

D:

The following turn includes a result linked by tool_use_id so the model can continue coherently; here, the controlling capability is that a structured error result communicates the failed operation and supports safe recovery, but returning an unrelated human message loses the execution association.

Question 333

A user may need a clear explanation when requested actions cannot be completed safely. What is the right implementation response?

  1. Write testable functional acceptance criteria
  2. Define fallback behavior under model outage
  3. Define the session continuity model
  4. Define accessible error and refusal behavior

Explanation

Correct Answer

D:

Usability criteria describe understandable errors, alternatives, and escalation paths; its operational boundary is that returning an empty model response leaves users unable to diagnose or recover.

Incorrect Answers

A:

Requirements should specify observable inputs, citation fields, and acceptance tests for the behavior; however, the critical mechanism here is that usability criteria describe understandable errors, alternatives, and escalation paths, and a statement that the system should be helpful is not independently verifiable.

B:

Availability requirements identify queueing, safe degraded responses, and recovery after service restoration; the requirement instead turns on this behavior: usability criteria describe understandable errors, alternatives, and escalation paths; blindly waiting for an API call does not guarantee a usable service.

C:

Session rules identify which prior context persists and when it must be invalidated; this cannot substitute for the behavior that usability criteria describe understandable errors, alternatives, and escalation paths, because accidentally reusing a cross-account history can expose sensitive context.

Question 334

A repository security audit should inspect risky files without filling the main assistant's context with raw logs. What is the most suitable pattern?

  1. Use a specialist subagent
  2. Use a single supervising agent
  3. Use a human approval checkpoint
  4. Use a summarizing worker for context compression

Explanation

Correct Answer

A:

An isolated worker handles a bounded analysis and returns a concise finding to the parent; the important constraint is that the subagent still needs a scoped tool set and cannot silently inherit unlimited production privileges.

Incorrect Answers

B:

A supervisor holds the overall objective and coordinates narrower workers through explicit handoffs; the requirement instead turns on this behavior: an isolated worker handles a bounded analysis and returns a concise finding to the parent; its instructions and budget must prevent cycles and duplicate work among workers.

C:

Execution pauses before the consequential action and resumes only after an authorized reviewer agrees; this cannot substitute for the behavior that an isolated worker handles a bounded analysis and returns a concise finding to the parent, because a natural-language promise to ask permission is weaker than an enforced approval gate.

D:

A dedicated summarizer condenses retained facts with provenance and open issues for the parent context; in contrast, an isolated worker handles a bounded analysis and returns a concise finding to the parent, while compression may omit nuance unless the system checks critical commitments and references.

Question 335

Several MCP-capable clients need the same governed repository search operations. What change is best aligned with the goal?

  1. Use a custom client tool for a private business operation
  2. Use a Skill for reusable procedures
  3. Use an MCP server for a reusable integration
  4. Combine MCP resources with a summarization prompt

Explanation

Correct Answer

C:

A protocol server exposes shared tools and resources with consistent discovery semantics; a crucial qualification is that it adds infrastructure and identity concerns compared with a single local function.

Incorrect Answers

A:

A user-defined tool schema plus local dispatcher safely bridges the model to that system; this addresses a separate concern, not the fact that a protocol server exposes shared tools and resources with consistent discovery semantics; the developer must operate the function and enforce its permissions.

B:

Skills package instructional workflows that an agent can reuse within supported clients; here, the controlling capability is that a protocol server exposes shared tools and resources with consistent discovery semantics, but they are guidance, not network APIs or execution permissions.

D:

Resources supply source data while a prompt governs how it should be analyzed; by comparison, the intended solution relies on this: a protocol server exposes shared tools and resources with consistent discovery semantics; neither component independently guarantees the summary is factually faithful.

Question 336

An assistant needs to execute short scripts submitted by outside users without access to production credentials. Which engineering practice is the strongest match?

  1. Apply a sandbox to untrusted generated code
  2. Use Anthropic-hosted managed execution where supported
  3. Apply a per-run timeout
  4. Run the harness in a controlled private environment

Explanation

Correct Answer

A:

Isolation constrains filesystem, processes, secrets, and network access independently of model reasoning; its operational boundary is that a prompt claiming code is safe cannot replace runtime containment.

Incorrect Answers

B:

Managed runtime can reduce infrastructure ownership for supported agent operations; this addresses a separate concern, not the fact that isolation constrains filesystem, processes, secrets, and network access independently of model reasoning; hosted deployment still requires review of data paths, available tools, and platform limitations.

C:

The host checks elapsed time and cancels or stops further tool iterations at the limit; here, the controlling capability is that isolation constrains filesystem, processes, secrets, and network access independently of model reasoning, but increasing token capacity does not impose a real execution deadline.

D:

A self-operated worker provides network and runtime control for executing the permitted functions; the missing capability is more specific: isolation constrains filesystem, processes, secrets, and network access independently of model reasoning; the operator must maintain patching, scaling, observability, and worker isolation.

Question 337

A browser request is abandoned while its backend is awaiting a slow model stream. Which measure directly addresses the problem?

  1. Record latency per integration stage
  2. Use async I/O for many concurrent inference requests
  3. Cancel timed-out tasks cleanly
  4. Bound concurrency with a semaphore

Explanation

Correct Answer

C:

Cancellation and timeout handling release resources and prevent orphaned work where supported, subject to this requirement: merely hiding the client spinner does not stop backend consumption.

Incorrect Answers

A:

Timed spans separate request phases and identify the bottleneck for targeted optimization; this cannot substitute for the behavior that cancellation and timeout handling release resources and prevent orphaned work where supported, because total elapsed time alone cannot localize the slow dependency.

B:

Asynchronous clients yield during I/O and resume when responses arrive; in contrast, cancellation and timeout handling release resources and prevent orphaned work where supported, while an event loop does not make a single model inference computationally faster.

D:

A concurrency limit caps simultaneous API work and prevents uncontrolled bursts; here, the controlling capability is that cancellation and timeout handling release resources and prevent orphaned work where supported, but launching every coroutine at once can trigger rate limiting and memory pressure.

Question 338

Intermittent 5xx responses persist across retries and the service must remain available. Which approach targets the deciding condition?

  1. Refresh authentication on unauthorized calls
  2. Back off on rate limiting
  3. Validate MIME type and size before upload
  4. Treat repeated server errors as a dependency outage

Explanation

Correct Answer

D:

A circuit breaker and fallback degrade gracefully while limiting repeated failed traffic; notably, a longer prompt or higher max_tokens does not restore an unavailable upstream server.

Incorrect Answers

A:

The operator checks key validity, workspace scope, and secret injection before retrying, whereas the requirement calls for another behavior: a circuit breaker and fallback degrade gracefully while limiting repeated failed traffic; sending the same invalid secret repeatedly will not solve the auth problem.

B:

Rate-limit handling respects retry hints and adds capped exponential delays with jitter; however, the critical mechanism here is that a circuit breaker and fallback degrade gracefully while limiting repeated failed traffic, and immediate repeated calls worsen overload and do not change the assigned quota.

C:

Server-side checks bound resource use and ensure supported input formats before model submission; the requirement instead turns on this behavior: a circuit breaker and fallback degrade gracefully while limiting repeated failed traffic; client-provided extensions and names are not reliable content validation alone.

Question 339

A model update starts generating unsafe tool calls and operations must be quickly curtailed. What is the most suitable pattern?

  1. Perform a least-privilege permissions review
  2. Canary a risky policy change
  3. Define a rollback and incident response path
  4. Use a policy-based refusal and safe redirect

Explanation

Correct Answer

C:

A response plan identifies feature shutdown, credential revocation, logs, and restoration steps; a crucial qualification is that there is no safe guarantee that tweaking the prompt alone fixes the incident.

Incorrect Answers

A:

Reviewing actual credentials and granted scopes shrinks the impact of a compromised prompt or tool; this addresses a separate concern, not the fact that a response plan identifies feature shutdown, credential revocation, logs, and restoration steps; the prompt's intended task does not limit the external credential automatically.

B:

A limited rollout measures refusal, error, and unauthorized action attempts before broad exposure; here, the controlling capability is that a response plan identifies feature shutdown, credential revocation, logs, and restoration steps, but immediate all-user changes make regressions harder to contain.

D:

A policy layer handles the request appropriately and communicates a safe boundary; by comparison, the intended solution relies on this: a response plan identifies feature shutdown, credential revocation, logs, and restoration steps; policy adherence should be evaluated; it is not guaranteed by a single warning prompt.

Question 340

An optimization once worked but token charges rise after a tool schema update. How should the implementation handle this need?

  1. Read cache diagnostics when hit rate falls
  2. Remove unnecessary cache breakpoints
  3. Include cache reads and writes in cost calculations
  4. Move unchanged long instructions into a shared prefix

Explanation

Correct Answer

A:

Cache usage metrics and prefix comparisons reveal the changed content causing misses, subject to this requirement: increasing the context window will not restore a mismatched cache prefix.

Incorrect Answers

B:

Simpler eligible breakpoint design reduces confusion and aligns with documented limits; here, the controlling capability is that cache usage metrics and prefix comparisons reveal the changed content causing misses, but arbitrary markers cannot create savings if prefixes keep changing.

C:

Cache creation and cache retrieval have distinct billing treatment that affects economics; the missing capability is more specific: cache usage metrics and prefix comparisons reveal the changed content causing misses; treating every token as an uncached input misstates true spend.

D:

Stable prefix design allows eligible prompt caching to amortize repeated input processing; by comparison, the intended solution relies on this: cache usage metrics and prefix comparisons reveal the changed content causing misses; putting request-specific content before a cache checkpoint defeats reuse.

Claude Certified Developer - Foundations (CCDV-F) Practice Tests