Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 47

Question 460

Two approval criteria are still open for a secure tool-execution boundary: the team must secure an agent against both malicious retrieved text and unsafe tool execution and require a person to approve an irreversible administrative change. Which two implementation actions satisfy them? Choose TWO.

  1. Apply an input-side policy gate and route flagged cases to a safe path.
  2. Validate or moderate final content before release.
  3. Pause execution for explicit approval at the high-impact action boundary.
  4. Combine prompt boundaries, scoped credentials, validation, and approval gates.
  5. Use layered policy enforcement and restrict powerful tools independently of chat text.

Explanation

Correct Answers

C:

To require a person to approve an irreversible administrative change as part of a design that must also secure an agent against both malicious retrieved text and unsafe tool execution, human approval is a policy-enforced transition, not a note written in the model's plan or a message displayed afterward; this is why the action is one of the two required selections.

D:

To secure an agent against both malicious retrieved text and unsafe tool execution as part of a design that must also require a person to approve an irreversible administrative change, defense in depth assumes individual filters can fail; independent enforcement boundaries reduce the chance of harmful side effects; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to identify prohibited requests before an automated agent invokes business tools because early filtering reduces inappropriate work, but later checks remain necessary because tool outputs can also introduce risks; the required outcomes are instead to secure an agent against both malicious retrieved text and unsafe tool execution and to require a person to approve an irreversible administrative change.

B:

This alternative serves the separate goal to prevent generated advice that violates the product's publication rules from being shown because output checks catch issues emerging during generation; they complement rather than replace input controls and tool restrictions; the required outcomes are instead to secure an agent against both malicious retrieved text and unsafe tool execution and to require a person to approve an irreversible administrative change.

E:

This alternative serves the separate goal to prevent an adversarial user from bypassing safety policy by roleplay instructions because adversarial prompting can target instruction hierarchy, so product controls and tool authorization must remain effective when wording is manipulated; the required outcomes are instead to secure an agent against both malicious retrieved text and unsafe tool execution and to require a person to approve an irreversible administrative change.

Question 461

Two approval criteria are still open for a long-context summarization workflow: the team must reduce the chance that critical instructions become buried in a huge prompt and keep only evidence that matters to the question when retrieval returns hundreds of documents. Which two implementation actions satisfy them? Choose TWO.

  1. Place essential constraints prominently and structure large evidence clearly.
  2. Revise one prompt component at a time against a regression evaluation set.
  3. Summarize prior decisions and preserve unresolved constraints.
  4. Rank and select relevant passages before building the model context.
  5. Reserve an output budget while enforcing input-context limits.

Explanation

Correct Answers

A:

To reduce the chance that critical instructions become buried in a huge prompt as part of a design that must also keep only evidence that matters to the question when retrieval returns hundreds of documents, ordering, headings, and explicit references help the model distinguish governing rules from supporting material in large requests; this is why the action is one of the two required selections.

D:

To keep only evidence that matters to the question when retrieval returns hundreds of documents as part of a design that must also reduce the chance that critical instructions become buried in a huge prompt, relevance filtering reduces distracting information and token usage; dumping every retrieved page can degrade grounding and exceed limits; this is why the action is one of the two required selections.

Incorrect Answers

B:

This alternative serves the separate goal to improve a prompt while preserving performance on previously solved cases because controlled prompt iteration distinguishes genuine quality gains from regressions or overfitting to one attractive example; the required outcomes are instead to reduce the chance that critical instructions become buried in a huge prompt and to keep only evidence that matters to the question when retrieval returns hundreds of documents.

C:

This alternative serves the separate goal to maintain the current task after older dialogue exceeds the active token budget because a faithful rolling summary carries forward objectives and commitments while avoiding a full-history replay on each call; the required outcomes are instead to reduce the chance that critical instructions become buried in a huge prompt and to keep only evidence that matters to the question when retrieval returns hundreds of documents.

E:

This alternative serves the separate goal to leave space for a model to generate a complete answer after adding source documents because input instructions, history, and evidence share the overall model capacity with generation, so filling it entirely can truncate output; the required outcomes are instead to reduce the chance that critical instructions become buried in a huge prompt and to keep only evidence that matters to the question when retrieval returns hundreds of documents.

Question 462

A release check on a Python Claude API backend found two unsatisfied conditions: the service must set consistent developer-controlled behavior separately from untrusted user text, and it must distinguish an end user's request from previous assistant turns in a dialogue. Which complete engineering plan satisfies both?

  1. Perform both steps: (1) Send instructions using the request's top-level system field; (2) read the completed stream's usage and final message metadata.
  2. Perform both steps: (1) Send instructions using the request's top-level system field; (2) assemble messages with correct user and assistant roles.
  3. Perform both steps: (1) Inspect response.content blocks and metadata through the SDK's typed representation; (2) read the completed stream's usage and final message metadata.
  4. Perform both steps: (1) Inspect response.content blocks and metadata through the SDK's typed representation; (2) assemble messages with correct user and assistant roles.

Explanation

Correct Answer

B:

The first change enables the team to set consistent developer-controlled behavior separately from untrusted user text because the system field separates higher-priority application guidance from user-message content and should not be simulated by ordinary quoted text; the second addresses the separate need to distinguish an end user's request from previous assistant turns in a dialogue because conversation structure influences how the model interprets dialogue and tool history; flattening roles into one text string loses that structure.

Incorrect Answers

A:

The first step is appropriate for the requirement to set consistent developer-controlled behavior separately from untrusted user text because the system field separates higher-priority application guidance from user-message content and should not be simulated by ordinary quoted text; the other step instead aims to account accurately for token consumption when a response is streamed because early text deltas do not necessarily carry authoritative final usage or stop information; accounting must wait for the appropriate terminal event, which does not meet the second requirement to distinguish an end user's request from previous assistant turns in a dialogue.

C:

Both steps target different results—avoid assuming a structured SDK response is a plain string and account accurately for token consumption when a response is streamed—because the SDK exposes block types, stop reason, and usage separately; naïve string concatenation can lose tool requests or other content and early text deltas do not necessarily carry authoritative final usage or stop information; accounting must wait for the appropriate terminal event; the required capabilities are instead to set consistent developer-controlled behavior separately from untrusted user text and to distinguish an end user's request from previous assistant turns in a dialogue.

D:

The second step meets the requirement to distinguish an end user's request from previous assistant turns in a dialogue because conversation structure influences how the model interprets dialogue and tool history; flattening roles into one text string loses that structure; the first step would only help to avoid assuming a structured SDK response is a plain string because the SDK exposes block types, stop reason, and usage separately; naïve string concatenation can lose tool requests or other content, leaving the separate requirement to set consistent developer-controlled behavior separately from untrusted user text unmet.

Question 463

A change request for a latency-sensitive model gateway combines two needs: keep a server responsive while several independent model calls wait on network I/O, and prevent a tool_use block from being displayed as if it were natural-language text. Which implementation pair addresses both, not just one?

  1. Perform both steps: (1) Compare both candidates on identical representative eval cases; (2) parse structured payloads with explicit failure handling.
  2. Perform both steps: (1) Compare both candidates on identical representative eval cases; (2) check each response block's type before extracting its fields.
  3. Perform both steps: (1) Use asynchronous client methods and limit in-flight calls; (2) parse structured payloads with explicit failure handling.
  4. Perform both steps: (1) Use asynchronous client methods and limit in-flight calls; (2) check each response block's type before extracting its fields.

Explanation

Correct Answer

D:

The first change enables the team to keep a server responsive while several independent model calls wait on network I/O because event-loop based I/O overlaps waits, but rate and resource bounds are still necessary to avoid overload; the second addresses the separate need to prevent a tool_use block from being displayed as if it were natural-language text because typed response objects can contain several block families; assuming every block has a text field causes logic errors.

Incorrect Answers

A:

Both steps target different results—choose between two supported model releases using evidence rather than reputation and reject incomplete or malformed response data before business logic consumes it—because holding inputs and scoring constant reveals quality and latency changes that generic model descriptions cannot settle and JSON parsing validates syntax, while separate schema and semantic checks verify types and permitted values; the required capabilities are instead to keep a server responsive while several independent model calls wait on network I/O and to prevent a tool_use block from being displayed as if it were natural-language text.

B:

The second step meets the requirement to prevent a tool_use block from being displayed as if it were natural-language text because typed response objects can contain several block families; assuming every block has a text field causes logic errors; the first step would only help to choose between two supported model releases using evidence rather than reputation because holding inputs and scoring constant reveals quality and latency changes that generic model descriptions cannot settle, leaving the separate requirement to keep a server responsive while several independent model calls wait on network I/O unmet.

C:

The first step is appropriate for the requirement to keep a server responsive while several independent model calls wait on network I/O because event-loop based I/O overlaps waits, but rate and resource bounds are still necessary to avoid overload; the other step instead aims to reject incomplete or malformed response data before business logic consumes it because JSON parsing validates syntax, while separate schema and semantic checks verify types and permitted values, which does not meet the second requirement to prevent a tool_use block from being displayed as if it were natural-language text.

Question 464

Two approval criteria are still open for a codebase used by several Claude Code operators: the team must make a reusable development procedure available to Claude Code with supporting files and introduce Claude Code to an unfamiliar codebase's conventions and commands. Which two implementation actions satisfy them? Choose TWO.

  1. Invoke the supported noninteractive or print-mode command interface.
  2. Configure a specialized Claude Code subagent with a narrow remit.
  3. Initialize project guidance and review the resulting CLAUDE.md content.
  4. Define policy in the appropriate managed or shared settings scope.
  5. Create a project skill with the documented skill metadata and content.

Explanation

Correct Answers

C:

To introduce Claude Code to an unfamiliar codebase's conventions and commands as part of a design that must also make a reusable development procedure available to Claude Code with supporting files, repository setup discovers useful project practices, but autogenerated guidance still needs a human check for accuracy and unnecessary detail; this is why the action is one of the two required selections.

E:

To make a reusable development procedure available to Claude Code with supporting files as part of a design that must also introduce Claude Code to an unfamiliar codebase's conventions and commands, a skill packages guidance or a workflow for reuse across relevant tasks; it is not identical to granting a runtime tool permission; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to run a Claude Code task from an automated CI job without an interactive terminal because headless operation outputs results for automation and requires explicit credential, permission, and failure handling in the CI environment; the required outcomes are instead to make a reusable development procedure available to Claude Code with supporting files and to introduce Claude Code to an unfamiliar codebase's conventions and commands.

B:

This alternative serves the separate goal to delegate an independent code review without filling the main coding session with details because subagents can focus on a task and return a summary, limiting noise in the parent session and allowing separate permissions when configured; the required outcomes are instead to make a reusable development procedure available to Claude Code with supporting files and to introduce Claude Code to an unfamiliar codebase's conventions and commands.

D:

This alternative serves the separate goal to enforce a team-approved permission policy despite local developer preferences because Claude Code has multiple configuration scopes; binding policy to the required scope matters more than placing a rule in an arbitrary JSON file; the required outcomes are instead to make a reusable development procedure available to Claude Code with supporting files and to introduce Claude Code to an unfamiliar codebase's conventions and commands.

Question 465

Two approval criteria are still open for a business workflow that calls Claude: the team must ensure each production feature can be tied to an approved acceptance test and detect rising inference errors after a feature reaches production. Which two implementation actions satisfy them? Choose TWO.

  1. Detect transport interruption and mark incomplete generated content.
  2. Enforce per-tenant token and request budgets in the backend.
  3. Maintain links between requirements, implementation changes, and verification cases.
  4. Keep versioned artifacts and a tested rollback procedure.
  5. Track latency, failure rates, token use, and output-quality signals over time.

Explanation

Correct Answers

C:

To ensure each production feature can be tied to an approved acceptance test as part of a design that must also detect rising inference errors after a feature reaches production, requirement traceability helps prevent a prompt or model update from silently changing behavior outside the agreed product scope; this is why the action is one of the two required selections.

E:

To detect rising inference errors after a feature reaches production as part of a design that must also ensure each production feature can be tied to an approved acceptance test, production health spans infrastructure and model behavior, so a healthy HTTP server alone does not prove the AI feature is functioning; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to avoid silently treating a broken streaming connection as a complete answer because network termination before a final message event can leave a partial response, which needs explicit recovery rather than success handling; the required outcomes are instead to ensure each production feature can be tied to an approved acceptance test and to detect rising inference errors after a feature reaches production.

B:

This alternative serves the separate goal to keep one customer from consuming an entire shared inference budget because application quotas isolate noisy tenants and control spending; provider-wide rate limits alone do not guarantee fairness between customers; the required outcomes are instead to ensure each production feature can be tied to an approved acceptance test and to detect rising inference errors after a feature reaches production.

D:

This alternative serves the separate goal to restore a known-working prompt and model configuration after quality degrades because model IDs, prompts, schemas, and tool contracts are release components; restoring just the application binary may not restore prior behavior; the required outcomes are instead to ensure each production feature can be tied to an approved acceptance test and to detect rising inference errors after a feature reaches production.

Question 466

For a model-family migration plan, the missing capability is to keep a server responsive while several independent model calls wait on network I/O. What should the developers implement?

  1. Use asynchronous client methods and limit in-flight calls.
  2. Run contract and integration tests against the upgraded SDK in staging.
  3. Route simple requests to a cheaper evaluated tier and escalate difficult ones.
  4. Iterate the SDK or HTTP event stream until its terminal event.

Explanation

Correct Answer

A:

To keep a server responsive while several independent model calls wait on network I/O in this model-family migration plan, event-loop based I/O overlaps waits, but rate and resource bounds are still necessary to avoid overload; the proposed implementation directly supplies that capability.

Incorrect Answers

B:

The proposed alternative helps to detect changed library behavior before rolling a new version into production because client-library version shifts can affect serialization or errors even without changes to business prompts, so they deserve release validation; that mechanism does not meet the different requirement to keep a server responsive while several independent model calls wait on network I/O.

C:

The proposed alternative helps to minimize average model cost when most requests are easy but some require deeper reasoning because cascading matches compute to task difficulty but requires a reliable escalation rule and measurement of misrouting; that mechanism does not meet the different requirement to keep a server responsive while several independent model calls wait on network I/O.

D:

The proposed alternative helps to process partial model output as it is delivered over a streaming connection because streaming clients must handle both content deltas and final status rather than assuming the first chunk is the complete response; that mechanism does not meet the different requirement to keep a server responsive while several independent model calls wait on network I/O.

Question 467

A design review for a tool-using incident triage agent asks for support to stop iterative tool use when a success condition or resource budget is reached, together with a way to send distinct request types to specialized handlers while keeping the routing rule inspectable. Select the two appropriate changes. Choose TWO.

  1. Classify the request and dispatch to an appropriate specialized workflow.
  2. Insert an approval gate before the high-risk action.
  3. Enforce explicit completion checks and a maximum iteration budget.
  4. Delegate source analysis to a subagent and return a concise result.
  5. Implement an application-owned Messages API agent loop.

Explanation

Correct Answers

A:

To send distinct request types to specialized handlers while keeping the routing rule inspectable as part of a design that must also stop iterative tool use when a success condition or resource budget is reached, routing separates distinct pathways without making the entire task autonomously planned; the classification decision must be observable and tested; this is why the action is one of the two required selections.

C:

To stop iterative tool use when a success condition or resource budget is reached as part of a design that must also send distinct request types to specialized handlers while keeping the routing rule inspectable, bounded loops avoid unending tool calls and uncontrolled spend while making incomplete work visible instead of silently continuing; this is why the action is one of the two required selections.

Incorrect Answers

B:

This alternative serves the separate goal to require human approval before an autonomous task crosses a high-impact decision boundary because human-in-the-loop control interrupts execution at a defined boundary; an agent's natural-language assurance cannot substitute for authorization; the required outcomes are instead to stop iterative tool use when a success condition or resource budget is reached and to send distinct request types to specialized handlers while keeping the routing rule inspectable.

D:

This alternative serves the separate goal to keep a large source-reading task from crowding the coordinator's working context because isolated workspaces let a specialist process extensive material while the parent receives only task-relevant evidence instead of every intermediate token; the required outcomes are instead to stop iterative tool use when a success condition or resource budget is reached and to send distinct request types to specialized handlers while keeping the routing rule inspectable.

E:

This alternative serves the separate goal to require precise control over message assembly, tool dispatch, and stopping behavior because a custom loop explicitly owns each model call, tool result, state update, and budget decision, trading flexibility for more engineering responsibility; the required outcomes are instead to stop iterative tool use when a success condition or resource budget is reached and to send distinct request types to specialized handlers while keeping the routing rule inspectable.

Question 468

In an agent access-control assessment, a release requirement is to reduce exposure after an API key is suspected compromised. Which implementation addresses that specific requirement?

  1. Treat retrieved content as untrusted and prevent its instructions from controlling tools.
  2. Apply an input-side policy gate and route flagged cases to a safe path.
  3. Redact secrets and protect log access with retention controls.
  4. Revoke and rotate the affected credential and update dependent services.

Explanation

Correct Answer

D:

To reduce exposure after an API key is suspected compromised in this agent access-control assessment, rotation restores credential control only when the old secret is disabled and consumers move to the replacement safely; the proposed implementation directly supplies that capability.

Incorrect Answers

A:

The proposed alternative helps to resist instructions embedded in a retrieved support ticket that demand secret disclosure because source material is data rather than authority; enforcing tool permissions and separating roles limits an attacker's ability to redirect execution; that mechanism does not meet the different requirement to reduce exposure after an API key is suspected compromised.

B:

The proposed alternative helps to identify prohibited requests before an automated agent invokes business tools because early filtering reduces inappropriate work, but later checks remain necessary because tool outputs can also introduce risks; that mechanism does not meet the different requirement to reduce exposure after an API key is suspected compromised.

C:

The proposed alternative helps to investigate user actions without leaking API keys into operational records because logs support accountability only when access and retention are restricted; copying all prompt content into logs creates another exposure surface; that mechanism does not meet the different requirement to reduce exposure after an API key is suspected compromised.

Question 469

A team maintaining a secure web application must meet 2 separate acceptance criteria: (1) keep a browser application from exposing the secret used for model calls; (2) restore a known-working prompt and model configuration after quality degrades. Which two implementation actions directly satisfy the stated requirements? Choose TWO.

  1. Route Claude requests through a trusted backend service.
  2. Attach the PDF using the supported document content form.
  3. Inspect response.content blocks and metadata through the SDK's typed representation.
  4. Keep versioned artifacts and a tested rollback procedure.
  5. Detect refusal and apply the product's allowed fallback or messaging policy.

Explanation

Correct Answers

A:

The test to keep a browser application from exposing the secret used for model calls is met because a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint; the remaining selection covers the separate need to restore a known-working prompt and model configuration after quality degrades.

D:

The test to restore a known-working prompt and model configuration after quality degrades is met because model IDs, prompts, schemas, and tool contracts are release components; restoring just the application binary may not restore prior behavior; the remaining selection covers the separate need to keep a browser application from exposing the secret used for model calls.

Incorrect Answers

B:

The action is useful for the separate capability to ask questions about information in a user-provided PDF through a supported document input because document analysis depends on actual supported content blocks and size limits; a local file path alone is not accessible to the remote model; however, the required test outcomes are to keep a browser application from exposing the secret used for model calls and to restore a known-working prompt and model configuration after quality degrades.

C:

The action is useful for the separate capability to avoid assuming a structured SDK response is a plain string because the SDK exposes block types, stop reason, and usage separately; naïve string concatenation can lose tool requests or other content; however, the required test outcomes are to keep a browser application from exposing the secret used for model calls and to restore a known-working prompt and model configuration after quality degrades.

E:

The action is useful for the separate capability to handle a safety refusal as a distinct application outcome because a refusal is not the same as a network failure; blindly retrying identical input may not produce an appropriate user experience; however, the required test outcomes are to keep a browser application from exposing the secret used for model calls and to restore a known-working prompt and model configuration after quality degrades.

Question 470

An end-to-end test for an operations research assistant checks whether it can send distinct request types to specialized handlers while keeping the routing rule inspectable and keep a large source-reading task from crowding the coordinator's working context. Which plan should be approved to satisfy both checks?

  1. Perform both steps: (1) Run independent branches concurrently and merge their outputs afterward; (2) delegate source analysis to a subagent and return a concise result.
  2. Perform both steps: (1) Classify the request and dispatch to an appropriate specialized workflow; (2) delegate source analysis to a subagent and return a concise result.
  3. Perform both steps: (1) Run independent branches concurrently and merge their outputs afterward; (2) enforce explicit completion checks and a maximum iteration budget.
  4. Perform both steps: (1) Classify the request and dispatch to an appropriate specialized workflow; (2) enforce explicit completion checks and a maximum iteration budget.

Explanation

Correct Answer

B:

The first change enables the team to send distinct request types to specialized handlers while keeping the routing rule inspectable because routing separates distinct pathways without making the entire task autonomously planned; the classification decision must be observable and tested; the second addresses the separate need to keep a large source-reading task from crowding the coordinator's working context because isolated workspaces let a specialist process extensive material while the parent receives only task-relevant evidence instead of every intermediate token.

Incorrect Answers

A:

The second step meets the requirement to keep a large source-reading task from crowding the coordinator's working context because isolated workspaces let a specialist process extensive material while the parent receives only task-relevant evidence instead of every intermediate token; the first step would only help to reduce elapsed time for independent extraction tasks that do not share mutable state because parallel orchestration shortens the critical path when calls have no dependencies, while the merger checks completeness before proceeding, leaving the separate requirement to send distinct request types to specialized handlers while keeping the routing rule inspectable unmet.

C:

Both steps target different results—reduce elapsed time for independent extraction tasks that do not share mutable state and stop iterative tool use when a success condition or resource budget is reached—because parallel orchestration shortens the critical path when calls have no dependencies, while the merger checks completeness before proceeding and bounded loops avoid unending tool calls and uncontrolled spend while making incomplete work visible instead of silently continuing; the required capabilities are instead to send distinct request types to specialized handlers while keeping the routing rule inspectable and to keep a large source-reading task from crowding the coordinator's working context.

D:

The first step is appropriate for the requirement to send distinct request types to specialized handlers while keeping the routing rule inspectable because routing separates distinct pathways without making the entire task autonomously planned; the classification decision must be observable and tested; the other step instead aims to stop iterative tool use when a success condition or resource budget is reached because bounded loops avoid unending tool calls and uncontrolled spend while making incomplete work visible instead of silently continuing, which does not meet the second requirement to keep a large source-reading task from crowding the coordinator's working context.

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