Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 77

Question 760

For a production inference application, the missing capability is to limit the blast radius of a new model configuration during rollout. What should the developers implement?

  1. Use the API's documented POST request with a JSON body.
  2. Exercise complete request-to-tool-to-response paths in a controlled environment.
  3. Handle stop_reason max_tokens as truncation and continue or revise the limit.
  4. Release to a small monitored cohort before broad deployment.

Explanation

Correct Answer

D:

To limit the blast radius of a new model configuration during rollout in this production inference application, canary exposure provides real traffic signals with an option to stop or roll back, unlike a simultaneous change for every customer; the proposed implementation directly supplies that capability.

Incorrect Answers

A:

The proposed alternative helps to send a model inference request with the operation expected by the REST endpoint because a state-changing inference operation is submitted as an HTTP request with an explicit method and payload, not inferred from a browser URL alone; that mechanism does not meet the different requirement to limit the blast radius of a new model configuration during rollout.

B:

The proposed alternative helps to catch mismatches between model calls, tool handlers, and application persistence because unit tests cannot reveal every serialization or permission issue across service boundaries; integration tests validate the connected contract; that mechanism does not meet the different requirement to limit the blast radius of a new model configuration during rollout.

C:

The proposed alternative helps to recover when a response ends because its generation budget was reached because a truncated completion may contain unfinished content, so treating it as a validated final answer risks losing required fields; that mechanism does not meet the different requirement to limit the blast radius of a new model configuration during rollout.

Question 761

Two approval criteria are still open for a business workflow that calls Claude: the team must ensure a rebuilt agent uses the same tested SDK dependency set and determine whether a quality regression came from a recent instruction edit. Which two implementation actions satisfy them? Choose TWO.

  1. Commit lockfiles and control library upgrades.
  2. Store prompt versions with release identifiers and rollback history.
  3. Release to a small monitored cohort before broad deployment.
  4. Require structured request IDs, tool traces, and version metadata with privacy controls.
  5. Use asynchronous SDK calls with bounded concurrency.

Explanation

Correct Answers

A:

To ensure a rebuilt agent uses the same tested SDK dependency set as part of a design that must also determine whether a quality regression came from a recent instruction edit, dependency pinning reduces environment drift; transitive SDK updates can change behavior even when prompt text is unchanged; this is why the action is one of the two required selections.

B:

To determine whether a quality regression came from a recent instruction edit as part of a design that must also ensure a rebuilt agent uses the same tested SDK dependency set, a prompt is a behavior-changing artifact; versioning makes comparisons and safe reversions possible during incident response; this is why the action is one of the two required selections.

Incorrect Answers

C:

This alternative serves the separate goal to limit the blast radius of a new model configuration during rollout because canary exposure provides real traffic signals with an option to stop or roll back, unlike a simultaneous change for every customer; the required outcomes are instead to ensure a rebuilt agent uses the same tested SDK dependency set and to determine whether a quality regression came from a recent instruction edit.

D:

This alternative serves the separate goal to enable an administrator to determine how a particular AI decision was produced because traceability depends on consistent operational evidence, not just retaining the final generated answer or a screenshot; the required outcomes are instead to ensure a rebuilt agent uses the same tested SDK dependency set and to determine whether a quality regression came from a recent instruction edit.

E:

This alternative serves the separate goal to serve many concurrent network-bound Claude requests without blocking worker threads unnecessarily because async I/O can overlap network waits, but it does not remove provider limits or make CPU-heavy postprocessing nonblocking; the required outcomes are instead to ensure a rebuilt agent uses the same tested SDK dependency set and to determine whether a quality regression came from a recent instruction edit.

Question 762

A team maintaining a cloud-hosted inference service must meet 3 separate acceptance criteria: (1) continue a chat with relevant prior turns despite the API being stateless; (2) stop a well-formed JSON response from introducing an impossible customer status; (3) have the server parse a JSON request body predictably. Which three implementation actions directly satisfy the stated requirements? Choose THREE.

  1. Include the necessary conversation history in each Messages API call.
  2. Keep versioned artifacts and a tested rollback procedure.
  3. Validate generated fields against domain rules after parsing.
  4. Require peer review focused on API, security, and tool-side effects.
  5. Assemble messages with correct user and assistant roles.
  6. Set the appropriate application/json content type and serialize a valid payload.

Explanation

Correct Answers

A:

The test to continue a chat with relevant prior turns despite the API being stateless is met because a response does not create server-side conversational memory by default; the client supplies retained turns and controls truncation; the remaining selections covers the separate need to stop a well-formed JSON response from introducing an impossible customer status and need to have the server parse a JSON request body predictably.

C:

The test to stop a well-formed JSON response from introducing an impossible customer status is met because a syntactically valid object may still contain invalid business values; server-side semantics determine whether it is acceptable; the remaining selections covers the separate need to continue a chat with relevant prior turns despite the API being stateless and need to have the server parse a JSON request body predictably.

F:

The test to have the server parse a JSON request body predictably is met because a JSON body requires both correct serialization and matching HTTP metadata; raw language-model text cannot replace a valid request object; the remaining selections covers the separate need to continue a chat with relevant prior turns despite the API being stateless and need to stop a well-formed JSON response from introducing an impossible customer status.

Incorrect Answers

B:

The action is useful for the separate capability 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; however, the required test outcomes are to continue a chat with relevant prior turns despite the API being stateless and to stop a well-formed JSON response from introducing an impossible customer status and to have the server parse a JSON request body predictably.

D:

The action is useful for the separate capability to catch unsafe credentials and incorrect tool calls before deployment because human review should check trust boundaries and external actions in addition to style; successful compilation cannot prove these behaviors are safe; however, the required test outcomes are to continue a chat with relevant prior turns despite the API being stateless and to stop a well-formed JSON response from introducing an impossible customer status and to have the server parse a JSON request body predictably.

E:

The action is useful for the separate capability 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; however, the required test outcomes are to continue a chat with relevant prior turns despite the API being stateless and to stop a well-formed JSON response from introducing an impossible customer status and to have the server parse a JSON request body predictably.

Question 763

An operational review of a hosted document-processing service reports distinct failures: it cannot distinguish API transport failures from a successfully returned model refusal, and it cannot cap the maximum output generated by a Messages API request. Which two repairs are required? Choose TWO.

  1. Inspect HTTP status and structured error payload before interpreting model output.
  2. Assemble messages with correct user and assistant roles.
  3. Include the necessary conversation history in each Messages API call.
  4. Configure the max_tokens request parameter appropriately.
  5. Include a typed image content block in the message.

Explanation

Correct Answers

A:

To distinguish API transport failures from a successfully returned model refusal as part of a design that must also cap the maximum output generated by a Messages API request, rate limits, authorization failures, and server errors happen at the request layer, unlike a normal response with a refusal stop reason; this is why the action is one of the two required selections.

D:

To cap the maximum output generated by a Messages API request as part of a design that must also distinguish API transport failures from a successfully returned model refusal, a generation limit controls maximum output length, not total cost or correctness; too low a value can truncate useful answers; this is why the action is one of the two required selections.

Incorrect Answers

B:

This alternative serves the separate goal 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 required outcomes are instead to distinguish API transport failures from a successfully returned model refusal and to cap the maximum output generated by a Messages API request.

C:

This alternative serves the separate goal to continue a chat with relevant prior turns despite the API being stateless because a response does not create server-side conversational memory by default; the client supplies retained turns and controls truncation; the required outcomes are instead to distinguish API transport failures from a successfully returned model refusal and to cap the maximum output generated by a Messages API request.

E:

This alternative serves the separate goal to ask Claude to interpret a supported image supplied as part of a request because multimodal input must use the API's image representation and supported sourcing; merely writing a filename in text does not transfer pixels; the required outcomes are instead to distinguish API transport failures from a successfully returned model refusal and to cap the maximum output generated by a Messages API request.

Question 764

A test for a TypeScript Claude integration fails because it cannot debug a failed request without exposing the user's confidential document. Which change fixes the identified failure?

  1. Version prompt files alongside relevant application code.
  2. Log status, timing, and correlation IDs with sensitive fields redacted.
  3. Refactor in bounded steps under existing tests.
  4. Require structured request IDs, tool traces, and version metadata with privacy controls.

Explanation

Correct Answer

B:

To debug a failed request without exposing the user's confidential document in this TypeScript Claude integration, operational records can identify a failure path without retaining raw secrets or protected prompt content; the proposed implementation directly supplies that capability.

Incorrect Answers

A:

The proposed alternative helps to trace exactly which prompt template was deployed with a code change because prompts change software behavior and should be reviewed and diffed like code, rather than edited anonymously in production; that mechanism does not meet the different requirement to debug a failed request without exposing the user's confidential document.

C:

The proposed alternative helps to preserve behavior while replacing a fragile parser in an AI integration because small changes constrain regression scope; a wholesale rewrite makes it harder to isolate whether model or parsing behavior changed; that mechanism does not meet the different requirement to debug a failed request without exposing the user's confidential document.

D:

The proposed alternative helps to enable an administrator to determine how a particular AI decision was produced because traceability depends on consistent operational evidence, not just retaining the final generated answer or a screenshot; that mechanism does not meet the different requirement to debug a failed request without exposing the user's confidential document.

Question 765

A planned upgrade to a multi-tenant support application has two requirements that must both pass: reproduce production behavior across deploys instead of silently adopting a new model release; compare candidate models before switching a mission-critical endpoint. Identify the two direct implementations. Choose TWO.

  1. Supply the required API version header.
  2. Treat stop_reason end_turn as a completed assistant turn.
  3. Document permitted processing regions and data-handling obligations before selecting a deployment.
  4. Pin a supported explicit model version where stability is required.
  5. Treat model migration as a tested deployment with staged traffic.

Explanation

Correct Answers

D:

To reproduce production behavior across deploys instead of silently adopting a new model release as part of a design that must also compare candidate models before switching a mission-critical endpoint, explicit version selection limits unexpected changes; a release process should evaluate upgrades and retirement timelines deliberately; this is why the action is one of the two required selections.

E:

To compare candidate models before switching a mission-critical endpoint as part of a design that must also reproduce production behavior across deploys instead of silently adopting a new model release, model upgrades may alter output distributions or supported features; staged validation reduces the chance of an unnoticed behavior shift; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to make a direct REST request with a declared supported API contract because the version header selects the protocol contract supported by the endpoint, preventing accidental dependence on unspecified behavior; the required outcomes are instead to reproduce production behavior across deploys instead of silently adopting a new model release and to compare candidate models before switching a mission-critical endpoint.

B:

This alternative serves the separate goal to detect that the assistant has naturally finished a response because the completion signal is explicit in the response metadata; it should not be inferred solely from a sentence ending in punctuation; the required outcomes are instead to reproduce production behavior across deploys instead of silently adopting a new model release and to compare candidate models before switching a mission-critical endpoint.

C:

This alternative serves the separate goal to keep sensitive case material within approved processing and retention boundaries because legal or organizational constraints must inform provider routing, logging, and retention rather than being added after data already flows; the required outcomes are instead to reproduce production behavior across deploys instead of silently adopting a new model release and to compare candidate models before switching a mission-critical endpoint.

Question 766

A release check on a Python Claude API backend found two unsatisfied conditions: the service must stop a well-formed JSON response from introducing an impossible customer status, and it must ensure refunds above a policy limit receive human sign-off. Which complete engineering plan satisfies both?

  1. Perform both steps: (1) Validate generated fields against domain rules after parsing; (2) enable required plugin or connector capabilities using least-privilege configuration.
  2. Perform both steps: (1) Queue or reject excess requests using bounded admission control; (2) enable required plugin or connector capabilities using least-privilege configuration.
  3. Perform both steps: (1) Validate generated fields against domain rules after parsing; (2) add application-owned approval logic around the refund tool.
  4. Perform both steps: (1) Queue or reject excess requests using bounded admission control; (2) add application-owned approval logic around the refund tool.

Explanation

Correct Answer

C:

The first change enables the team to stop a well-formed JSON response from introducing an impossible customer status because a syntactically valid object may still contain invalid business values; server-side semantics determine whether it is acceptable; the second addresses the separate need to ensure refunds above a policy limit receive human sign-off because decisions that create financial side effects require deterministic authorization checks beyond a model's recommendation.

Incorrect Answers

A:

The first step is appropriate for the requirement to stop a well-formed JSON response from introducing an impossible customer status because a syntactically valid object may still contain invalid business values; server-side semantics determine whether it is acceptable; the other step instead aims to expose external business tools only to workflows that actually need them because tools expand the application's action surface; indiscriminate activation increases risk and can confuse model selection among similar functions, which does not meet the second requirement to ensure refunds above a policy limit receive human sign-off.

B:

Both steps target different results—shield customer traffic when inference capacity is exhausted and expose external business tools only to workflows that actually need them—because controlled queuing protects system latency and prevents unbounded retries from overwhelming downstream provider quotas and tools expand the application's action surface; indiscriminate activation increases risk and can confuse model selection among similar functions; the required capabilities are instead to stop a well-formed JSON response from introducing an impossible customer status and to ensure refunds above a policy limit receive human sign-off.

D:

The second step meets the requirement to ensure refunds above a policy limit receive human sign-off because decisions that create financial side effects require deterministic authorization checks beyond a model's recommendation; the first step would only help to shield customer traffic when inference capacity is exhausted because controlled queuing protects system latency and prevents unbounded retries from overwhelming downstream provider quotas, leaving the separate requirement to stop a well-formed JSON response from introducing an impossible customer status unmet.

Question 767

For a customer-facing AI feature, a design is acceptable only if it can detect that the assistant has naturally finished a response while also being able to account accurately for token consumption when a response is streamed. Which paired implementation is sufficient?

  1. Perform both steps: (1) Treat stop_reason end_turn as a completed assistant turn; (2) enable streaming and consume server-sent events in order.
  2. Perform both steps: (1) Configure the max_tokens request parameter appropriately; (2) enable streaming and consume server-sent events in order.
  3. Perform both steps: (1) Configure the max_tokens request parameter appropriately; (2) read the completed stream's usage and final message metadata.
  4. Perform both steps: (1) Treat stop_reason end_turn as a completed assistant turn; (2) read the completed stream's usage and final message metadata.

Explanation

Correct Answer

D:

The first change enables the team to detect that the assistant has naturally finished a response because the completion signal is explicit in the response metadata; it should not be inferred solely from a sentence ending in punctuation; the second addresses the separate need 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.

Incorrect Answers

A:

The first step is appropriate for the requirement to detect that the assistant has naturally finished a response because the completion signal is explicit in the response metadata; it should not be inferred solely from a sentence ending in punctuation; the other step instead aims to show generated text progressively instead of waiting for the full response because streaming transmits partial deltas and completion events, allowing earlier display while requiring the client to assemble content safely, which does not meet the second requirement to account accurately for token consumption when a response is streamed.

B:

Both steps target different results—cap the maximum output generated by a Messages API request and show generated text progressively instead of waiting for the full response—because a generation limit controls maximum output length, not total cost or correctness; too low a value can truncate useful answers and streaming transmits partial deltas and completion events, allowing earlier display while requiring the client to assemble content safely; the required capabilities are instead to detect that the assistant has naturally finished a response and to account accurately for token consumption when a response is streamed.

C:

The second step meets the requirement 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; the first step would only help to cap the maximum output generated by a Messages API request because a generation limit controls maximum output length, not total cost or correctness; too low a value can truncate useful answers, leaving the separate requirement to detect that the assistant has naturally finished a response unmet.

Question 768

Two acceptance tests block a production inference application: one requires the ability to expose external business tools only to workflows that actually need them; the other requires the ability to prevent a retrieved article's embedded commands from being treated as developer instructions. Which plan clears both tests?

  1. Perform both steps: (1) Route Claude requests through a trusted backend service; (2) validate generated fields against domain rules after parsing.
  2. Perform both steps: (1) Enable required plugin or connector capabilities using least-privilege configuration; (2) validate generated fields against domain rules after parsing.
  3. Perform both steps: (1) Route Claude requests through a trusted backend service; (2) delimit retrieved text as untrusted source material.
  4. Perform both steps: (1) Enable required plugin or connector capabilities using least-privilege configuration; (2) delimit retrieved text as untrusted source material.

Explanation

Correct Answer

D:

The first change enables the team to expose external business tools only to workflows that actually need them because tools expand the application's action surface; indiscriminate activation increases risk and can confuse model selection among similar functions; the second addresses the separate need to prevent a retrieved article's embedded commands from being treated as developer instructions because explicit trust boundaries help the model interpret documents as evidence rather than operational orders, and runtime controls must enforce privileged actions.

Incorrect Answers

A:

Both steps target different results—keep a browser application from exposing the secret used for model calls and stop a well-formed JSON response from introducing an impossible customer status—because a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint and a syntactically valid object may still contain invalid business values; server-side semantics determine whether it is acceptable; the required capabilities are instead to expose external business tools only to workflows that actually need them and to prevent a retrieved article's embedded commands from being treated as developer instructions.

B:

The first step is appropriate for the requirement to expose external business tools only to workflows that actually need them because tools expand the application's action surface; indiscriminate activation increases risk and can confuse model selection among similar functions; the other step instead aims to stop a well-formed JSON response from introducing an impossible customer status because a syntactically valid object may still contain invalid business values; server-side semantics determine whether it is acceptable, which does not meet the second requirement to prevent a retrieved article's embedded commands from being treated as developer instructions.

C:

The second step meets the requirement to prevent a retrieved article's embedded commands from being treated as developer instructions because explicit trust boundaries help the model interpret documents as evidence rather than operational orders, and runtime controls must enforce privileged actions; the first step would only help to keep a browser application from exposing the secret used for model calls because a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint, leaving the separate requirement to expose external business tools only to workflows that actually need them unmet.

Question 769

The technical owner of a business workflow that calls Claude has to customize one developer's default Claude Code behavior across local projects. A separate requirement is to avoid a plugin update unexpectedly changing tools available to a coding agent. Which two-part remediation covers both?

  1. Perform both steps: (1) Use the user's own Claude Code settings file; (2) track provider deprecation notices and test migration before the removal date.
  2. Perform both steps: (1) Detect transport interruption and mark incomplete generated content; (2) track provider deprecation notices and test migration before the removal date.
  3. Perform both steps: (1) Detect transport interruption and mark incomplete generated content; (2) track enabled plugins and their compatible versions as release dependencies.
  4. Perform both steps: (1) Use the user's own Claude Code settings file; (2) track enabled plugins and their compatible versions as release dependencies.

Explanation

Correct Answer

D:

The first change enables the team to customize one developer's default Claude Code behavior across local projects because user settings are scoped to one person, so they should not be relied on for enforcing team-wide policy; the second addresses the separate need to avoid a plugin update unexpectedly changing tools available to a coding agent because plugin capability changes can affect tool names and permissions, so production configuration should be reproducible and reviewed.

Incorrect Answers

A:

The first step is appropriate for the requirement to customize one developer's default Claude Code behavior across local projects because user settings are scoped to one person, so they should not be relied on for enforcing team-wide policy; the other step instead aims to avoid an outage when a pinned model version approaches retirement because lifecycle management treats model availability as a dependency and validates output regressions before switching production traffic, which does not meet the second requirement to avoid a plugin update unexpectedly changing tools available to a coding agent.

B:

Both steps target different results—avoid silently treating a broken streaming connection as a complete answer and avoid an outage when a pinned model version approaches retirement—because network termination before a final message event can leave a partial response, which needs explicit recovery rather than success handling and lifecycle management treats model availability as a dependency and validates output regressions before switching production traffic; the required capabilities are instead to customize one developer's default Claude Code behavior across local projects and to avoid a plugin update unexpectedly changing tools available to a coding agent.

C:

The second step meets the requirement to avoid a plugin update unexpectedly changing tools available to a coding agent because plugin capability changes can affect tool names and permissions, so production configuration should be reproducible and reviewed; the first step would only help 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, leaving the separate requirement to customize one developer's default Claude Code behavior across local projects unmet.

Question 770

A design review for a high-volume Messages API gateway asks for support to support Claude models accessed through different supported cloud offerings, together with a way to handle a safety refusal as a distinct application outcome. Select the two appropriate changes. Choose TWO.

  1. Detect transport interruption and mark incomplete generated content.
  2. Configure an available server-side tool with the required API definition.
  3. Detect refusal and apply the product's allowed fallback or messaging policy.
  4. Use the provider-specific endpoint, authentication, and request conventions.
  5. Include a typed image content block in the message.

Explanation

Correct Answers

C:

To handle a safety refusal as a distinct application outcome as part of a design that must also support Claude models accessed through different supported cloud offerings, a refusal is not the same as a network failure; blindly retrying identical input may not produce an appropriate user experience; this is why the action is one of the two required selections.

D:

To support Claude models accessed through different supported cloud offerings as part of a design that must also handle a safety refusal as a distinct application outcome, Claude API, Bedrock, and Vertex-style integrations are not interchangeable URLs; host authentication and service wrappers differ; 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 support Claude models accessed through different supported cloud offerings and to handle a safety refusal as a distinct application outcome.

B:

This alternative serves the separate goal to use provider-hosted search or execution without implementing the tool body locally because provider-executed tools are handled in Anthropic's environment and differ from client functions that the caller must run; the required outcomes are instead to support Claude models accessed through different supported cloud offerings and to handle a safety refusal as a distinct application outcome.

E:

This alternative serves the separate goal to ask Claude to interpret a supported image supplied as part of a request because multimodal input must use the API's image representation and supported sourcing; merely writing a filename in text does not transfer pixels; the required outcomes are instead to support Claude models accessed through different supported cloud offerings and to handle a safety refusal as a distinct application outcome.

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