Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 38

Question 370

The acceptance criteria for an agent platform change state that it must retain complete control over a small set of proprietary tool calls and their execution order. Which implementation aligns with that criterion?

  1. Execute independent requested tools concurrently
  2. Implement a custom Messages API tool loop
  3. Use the Claude Agent SDK for an integrated tool-driven agent
  4. Deploy agent workers close to controlled data sources

Explanation

Correct Answer

B:

The application receives tool requests, validates inputs, executes authorized functions, and sends tool results back; this behavior directly supports the need to retain complete control over a small set of proprietary tool calls and their execution order. The Messages API does not automatically execute arbitrary client tools.

Incorrect Answers

A:

When an application needs to reduce turnaround when an agent asks for unrelated reads in the same response, parallel execution is safe for independent operations while every result must still be matched to its call ID; the current capability must instead retain complete control over a small set of proprietary tool calls and their execution order, while mutating tools with data dependencies should not be blindly parallelized.

C:

This approach serves the separate need to build a coding agent that needs managed tool execution, sessions, and event handling because the SDK provides agent-loop primitives and integrations that remove much of the custom harness work; it does not establish the outcome to retain complete control over a small set of proprietary tool calls and their execution order, and it still requires explicit permissions, safety policies, and operational boundaries.

D:

A self-hosted harness can enforce infrastructure, network, and retention controls for the surrounding tools; despite being useful when a service must meet a requirement that proprietary repository contents remain within an approved execution environment, it is not a solution to retain complete control over a small set of proprietary tool calls and their execution order; hosted orchestration does not automatically inherit a company's private-network policies.

Question 371

In an inference routing project, acceptance test A checks whether a component can estimate whether a multilingual request will fit a model context window. Acceptance test B checks whether a different component can classify a familiar support message where labels and criteria are already clear. Which plan satisfies both tests?

  1. First component: Reduce reasoning effort on trivial routing. Second component: Treat model outputs as stochastic.
  2. First component: Budget tokens rather than characters. Second component: Reduce reasoning effort on trivial routing.
  3. First component: Budget tokens rather than characters. Second component: Apply zero-shot instructions for a straightforward task.
  4. First component: Treat model outputs as stochastic. Second component: Apply zero-shot instructions for a straightforward task.

Explanation

Correct Answer

C:

The first proposal fits its assigned role because tokenization splits text into model-specific units, so character counts are not a reliable context proxy, while the second proposal fits its assigned role because direct instructions can be sufficient when the task is simple and well specified; their respective acceptance criteria are to estimate whether a multilingual request will fit a model context window and to classify a familiar support message where labels and criteria are already clear.

Incorrect Answers

A:

The first proposal serves the different purpose to shorten latency and spend on tasks that need only basic categorization instead of the assigned first role because lower effort can suit predictable work where deeper reasoning adds little measurable quality, while the second proposal serves the different purpose to design tests that tolerate benign wording variation while detecting semantic failures instead of the assigned second role because non-determinism requires evaluation by criteria and repeated examples rather than exact-string equality everywhere; their respective acceptance criteria are to estimate whether a multilingual request will fit a model context window and to classify a familiar support message where labels and criteria are already clear.

B:

The first proposal fits its assigned role because tokenization splits text into model-specific units, so character counts are not a reliable context proxy, while the second proposal serves the different purpose to shorten latency and spend on tasks that need only basic categorization instead of the assigned second role because lower effort can suit predictable work where deeper reasoning adds little measurable quality; their respective acceptance criteria are to estimate whether a multilingual request will fit a model context window and to classify a familiar support message where labels and criteria are already clear.

D:

The first proposal serves the different purpose to design tests that tolerate benign wording variation while detecting semantic failures instead of the assigned first role because non-determinism requires evaluation by criteria and repeated examples rather than exact-string equality everywhere, while the second proposal fits its assigned role because direct instructions can be sufficient when the task is simple and well specified; their respective acceptance criteria are to estimate whether a multilingual request will fit a model context window and to classify a familiar support message where labels and criteria are already clear.

Question 372

Two features of an operations orchestration system are under review. One feature must choose the right specialised handling path from the request category; the other must bound the runtime of an agent that repeatedly searches without converging. Which plan matches the respective needs?

  1. First component: Apply routing before task execution. Second component: Use parallel fan-out and a reconciliation step.
  2. First component: Keep a single agent with a small tool set. Second component: Use an explicit completion condition and step budget.
  3. First component: Apply routing before task execution. Second component: Use an explicit completion condition and step budget.
  4. First component: Use parallel fan-out and a reconciliation step. Second component: Keep a single agent with a small tool set.

Explanation

Correct Answer

C:

The first proposal fits its assigned role because an inexpensive classification or rule-based router sends each request to the applicable predefined workflow, while the second proposal fits its assigned role because a maximum iteration or cost budget plus a clear success criterion stops runaway tool loops; their respective acceptance criteria are to choose the right specialised handling path from the request category and to bound the runtime of an agent that repeatedly searches without converging.

Incorrect Answers

A:

The first proposal fits its assigned role because an inexpensive classification or rule-based router sends each request to the applicable predefined workflow, while the second proposal serves the different purpose to analyze several independent document sections concurrently and combine the results instead of the assigned second role because parallel tasks reduce critical-path latency when they do not depend on one another; a join step resolves conflicting findings; their respective acceptance criteria are to choose the right specialised handling path from the request category and to bound the runtime of an agent that repeatedly searches without converging.

B:

The first proposal serves the different purpose to answer routine account questions without the overhead of orchestrating multiple agents instead of the assigned first role because a single constrained executor simplifies state, authentication, logging, and latency where specialization adds little, while the second proposal fits its assigned role because a maximum iteration or cost budget plus a clear success criterion stops runaway tool loops; their respective acceptance criteria are to choose the right specialised handling path from the request category and to bound the runtime of an agent that repeatedly searches without converging.

D:

The first proposal serves the different purpose to analyze several independent document sections concurrently and combine the results instead of the assigned first role because parallel tasks reduce critical-path latency when they do not depend on one another; a join step resolves conflicting findings, while the second proposal serves the different purpose to answer routine account questions without the overhead of orchestrating multiple agents instead of the assigned second role because a single constrained executor simplifies state, authentication, logging, and latency where specialization adds little; their respective acceptance criteria are to choose the right specialised handling path from the request category and to bound the runtime of an agent that repeatedly searches without converging.

Question 373

During an external system access rollout, two different failure paths need remediation. The first component must share a repeatable analyst workflow across supported clients; separately, the second must prevent a server update from returning incompatible payloads to clients. Which changes should be adopted?

  1. First component: Expose an MCP prompt template. Second component: Expose an MCP resource for reference content.
  2. First component: Expose an MCP prompt template. Second component: Validate MCP tool schemas and results.
  3. First component: Use stdio for a local MCP process. Second component: Validate MCP tool schemas and results.
  4. First component: Expose an MCP resource for reference content. Second component: Use stdio for a local MCP process.

Explanation

Correct Answer

B:

The first proposal fits its assigned role because prompts describe reusable message templates or workflows offered by the server, while the second proposal fits its assigned role because contract tests check declared capabilities, argument shapes, and expected response content; their respective acceptance criteria are to share a repeatable analyst workflow across supported clients and to prevent a server update from returning incompatible payloads to clients.

Incorrect Answers

A:

The first proposal fits its assigned role because prompts describe reusable message templates or workflows offered by the server, while the second proposal serves the different purpose to make a versioned product catalog available for read-oriented context retrieval instead of the assigned second role because resources provide identifiable data that clients can fetch for model context; their respective acceptance criteria are to share a repeatable analyst workflow across supported clients and to prevent a server update from returning incompatible payloads to clients.

C:

The first proposal serves the different purpose to connect a desktop development tool to a trusted local server process instead of the assigned first role because standard input and output transport is suitable for a client-managed local subprocess, while the second proposal fits its assigned role because contract tests check declared capabilities, argument shapes, and expected response content; their respective acceptance criteria are to share a repeatable analyst workflow across supported clients and to prevent a server update from returning incompatible payloads to clients.

D:

The first proposal serves the different purpose to make a versioned product catalog available for read-oriented context retrieval instead of the assigned first role because resources provide identifiable data that clients can fetch for model context, while the second proposal serves the different purpose to connect a desktop development tool to a trusted local server process instead of the assigned second role because standard input and output transport is suitable for a client-managed local subprocess; their respective acceptance criteria are to share a repeatable analyst workflow across supported clients and to prevent a server update from returning incompatible payloads to clients.

Question 374

The owner of a prompt engineering integration wants to let the model summarize a hostile email without obeying its embedded commands. Which proposal provides the required mechanism?

  1. Delimit untrusted source material
  2. Specify output format in the prompt
  3. Request evidence before conclusions
  4. State one measurable outcome per instruction
  5. Iterate prompts against an evaluation set

Explanation

Correct Answer

A:

Clear source boundaries and explicit treatment as quoted evidence help reduce instruction confusion; this behavior directly supports the need to let the model summarize a hostile email without obeying its embedded commands. Delimiters alone cannot replace permission checks on dangerous tools.

Incorrect Answers

B:

For the alternate objective to obtain named fields in a consistent response layout for subsequent processing, explicit field descriptions guide generation toward the consumer's expected structure; the feature here must let the model summarize a hostile email without obeying its embedded commands, and formatting guidance still requires machine validation at the boundary.

C:

Asking for citations to provided excerpts promotes inspectable grounding and makes verification possible; this applies to the different goal to reduce unsupported claims in answers about internal procedures, not the requirement to let the model summarize a hostile email without obeying its embedded commands; a fabricated citation can still look plausible without source checking.

D:

When an application needs to stop a summarizer from producing unrelated analysis instead of required action items, clear task goals and acceptance criteria reduce competing interpretations; the current capability must instead let the model summarize a hostile email without obeying its embedded commands, while a vague request for a good summary provides no testable target.

E:

Controlled revisions measured on stable examples show whether a change actually helps; the supported outcome is to improve extraction quality after discovering consistent edge-case failures, whereas this scenario requires the ability to let the model summarize a hostile email without obeying its embedded commands; optimizing against a single anecdote can degrade other cases.

Question 375

Two separate tickets in a retrieval pipeline backlog require different controls. One needs to prevent outdated user preferences from being treated as current task instructions; the other needs to keep an embedded webpage instruction from becoming higher-priority application policy. Which implementation pair is correct?

  1. First component: Separate long-term memory from active scratchpad. Second component: Refresh stale evidence before reuse.
  2. First component: Refresh stale evidence before reuse. Second component: Compact completed task history.
  3. First component: Separate long-term memory from active scratchpad. Second component: Preserve instruction and evidence boundaries.
  4. First component: Compact completed task history. Second component: Preserve instruction and evidence boundaries.

Explanation

Correct Answer

C:

The first proposal fits its assigned role because explicit storage and freshness rules govern what is retained and when it is reintroduced, while the second proposal fits its assigned role because context assembly must label external text as data and avoid granting it authority over tool use; their respective acceptance criteria are to prevent outdated user preferences from being treated as current task instructions and to keep an embedded webpage instruction from becoming higher-priority application policy.

Incorrect Answers

A:

The first proposal fits its assigned role because explicit storage and freshness rules govern what is retained and when it is reintroduced, while the second proposal serves the different purpose to avoid answering a pricing question from last month's retrieved cache instead of the assigned second role because freshness checks determine whether a saved passage remains authoritative for a dynamic fact; their respective acceptance criteria are to prevent outdated user preferences from being treated as current task instructions and to keep an embedded webpage instruction from becoming higher-priority application policy.

B:

The first proposal serves the different purpose to avoid answering a pricing question from last month's retrieved cache instead of the assigned first role because freshness checks determine whether a saved passage remains authoritative for a dynamic fact, while the second proposal serves the different purpose to let an assistant continue a long incident investigation without losing decisions instead of the assigned second role because a structured summary preserves established facts, pending work, and unresolved assumptions in fewer tokens; their respective acceptance criteria are to prevent outdated user preferences from being treated as current task instructions and to keep an embedded webpage instruction from becoming higher-priority application policy.

D:

The first proposal serves the different purpose to let an assistant continue a long incident investigation without losing decisions instead of the assigned first role because a structured summary preserves established facts, pending work, and unresolved assumptions in fewer tokens, while the second proposal fits its assigned role because context assembly must label external text as data and avoid granting it authority over tool use; their respective acceptance criteria are to prevent outdated user preferences from being treated as current task instructions and to keep an embedded webpage instruction from becoming higher-priority application policy.

Question 376

Two features of an inference routing system are under review. One feature must detect behavior changes before a newly released model reaches customers; the other must select the right model for an application with tool and network overhead. Which plan matches the respective needs?

  1. First component: Prefer a Sonnet-class model for balanced workloads. Second component: Run task-specific model evaluations.
  2. First component: Run task-specific model evaluations. Second component: Measure end-to-end latency, not just token speed.
  3. First component: Pin model versions during a migration. Second component: Measure end-to-end latency, not just token speed.
  4. First component: Pin model versions during a migration. Second component: Prefer a Sonnet-class model for balanced workloads.

Explanation

Correct Answer

C:

The first proposal fits its assigned role because explicit IDs and a staged rollout make differences attributable and reversible, while the second proposal fits its assigned role because total response time includes queuing, input processing, generation, and downstream integration; their respective acceptance criteria are to detect behavior changes before a newly released model reaches customers and to select the right model for an application with tool and network overhead.

Incorrect Answers

A:

The first proposal serves the different purpose to serve a mixed coding and support assistant at moderate latency and cost instead of the assigned first role because the balanced tier often offers a practical quality-throughput compromise, subject to task-level testing, while the second proposal serves the different purpose to choose between releases using an actual organization's error and latency measurements instead of the assigned second role because representative tests expose differences hidden by generic capability labels; their respective acceptance criteria are to detect behavior changes before a newly released model reaches customers and to select the right model for an application with tool and network overhead.

B:

The first proposal serves the different purpose to choose between releases using an actual organization's error and latency measurements instead of the assigned first role because representative tests expose differences hidden by generic capability labels, while the second proposal fits its assigned role because total response time includes queuing, input processing, generation, and downstream integration; their respective acceptance criteria are to detect behavior changes before a newly released model reaches customers and to select the right model for an application with tool and network overhead.

D:

The first proposal fits its assigned role because explicit IDs and a staged rollout make differences attributable and reversible, while the second proposal serves the different purpose to serve a mixed coding and support assistant at moderate latency and cost instead of the assigned second role because the balanced tier often offers a practical quality-throughput compromise, subject to task-level testing; their respective acceptance criteria are to detect behavior changes before a newly released model reaches customers and to select the right model for an application with tool and network overhead.

Question 377

A release plan for an inference routing platform has two acceptance checks: component one must avoid rejecting tool schemas because serialized values are invalid; component two must avoid treating a tool call or thinking block as user-visible answer text. Which proposal passes both?

  1. First component: Use JSON-safe types in request bodies. Second component: Send the correct API authentication header.
  2. First component: Send the correct API authentication header. Second component: Use direct HTTP REST integration for unsupported runtimes.
  3. First component: Use direct HTTP REST integration for unsupported runtimes. Second component: Parse responses by content block type.
  4. First component: Use JSON-safe types in request bodies. Second component: Parse responses by content block type.

Explanation

Correct Answer

D:

The first proposal fits its assigned role because structured request encoding enforces objects, arrays, strings, booleans, and numeric representations, while the second proposal fits its assigned role because structured outputs can contain different block kinds that need type-aware handling; their respective acceptance criteria are to avoid rejecting tool schemas because serialized values are invalid and to avoid treating a tool call or thinking block as user-visible answer text.

Incorrect Answers

A:

The first proposal fits its assigned role because structured request encoding enforces objects, arrays, strings, booleans, and numeric representations, while the second proposal serves the different purpose to avoid unauthenticated model requests from a backend worker instead of the assigned second role because the API validates the provided credential before accepting a request; their respective acceptance criteria are to avoid rejecting tool schemas because serialized values are invalid and to avoid treating a tool call or thinking block as user-visible answer text.

B:

The first proposal serves the different purpose to avoid unauthenticated model requests from a backend worker instead of the assigned first role because the API validates the provided credential before accepting a request, while the second proposal serves the different purpose to call the Messages endpoint from a language without a mature provider SDK instead of the assigned second role because a standards-based HTTP client can send JSON requests with the correct headers and decode responses; their respective acceptance criteria are to avoid rejecting tool schemas because serialized values are invalid and to avoid treating a tool call or thinking block as user-visible answer text.

C:

The first proposal serves the different purpose to call the Messages endpoint from a language without a mature provider SDK instead of the assigned first role because a standards-based HTTP client can send JSON requests with the correct headers and decode responses, while the second proposal fits its assigned role because structured outputs can contain different block kinds that need type-aware handling; their respective acceptance criteria are to avoid rejecting tool schemas because serialized values are invalid and to avoid treating a tool call or thinking block as user-visible answer text.

Question 378

An operations orchestration group is preparing a production change intended to prevent an autonomous investigation from issuing irreversible account changes without review. Which technique matches that intent?

  1. Use an explicit completion condition and step budget
  2. Adopt a supervisor with specialist workers
  3. Introduce a human approval checkpoint
  4. Keep a single agent with a small tool set

Explanation

Correct Answer

C:

A human gate separates analysis from externally consequential action and gives a reviewer a veto; this behavior directly supports the need to prevent an autonomous investigation from issuing irreversible account changes without review. A system-prompt request to be careful is not an enforceable approval workflow.

Incorrect Answers

A:

When an application needs to bound the runtime of an agent that repeatedly searches without converging, a maximum iteration or cost budget plus a clear success criterion stops runaway tool loops; the current capability must instead prevent an autonomous investigation from issuing irreversible account changes without review, while merely asking the model to finish quickly does not enforce execution limits.

B:

A managing agent delegates bounded subtasks to specialists and reconciles their outputs against a common objective; the supported outcome is to coordinate separate billing, compliance, and engineering analyses under one final decision, whereas this scenario requires the ability to prevent an autonomous investigation from issuing irreversible account changes without review; independent workers with no coordinator can return inconsistent recommendations.

D:

A single constrained executor simplifies state, authentication, logging, and latency where specialization adds little; despite being useful when a service must answer routine account questions without the overhead of orchestrating multiple agents, it is not a solution to prevent an autonomous investigation from issuing irreversible account changes without review; a multi-agent hierarchy is not a default requirement for tool use.

Question 379

During an agent extensibility rollout, two different failure paths need remediation. The first component must require stricter approval for actions that change customer data; separately, the second must avoid creating duplicate tickets after a tool call times out. Which changes should be adopted?

  1. First component: Separate read tools from write tools. Second component: Use idempotency for retried mutations.
  2. First component: Separate read tools from write tools. Second component: Map tool_use IDs to results.
  3. First component: Restrict tool network credentials. Second component: Use idempotency for retried mutations.
  4. First component: Map tool_use IDs to results. Second component: Restrict tool network credentials.

Explanation

Correct Answer

A:

The first proposal fits its assigned role because distinct interfaces simplify least-privilege grants and side-effect controls, while the second proposal fits its assigned role because stable operation identifiers let downstream systems detect repeated submissions; their respective acceptance criteria are to require stricter approval for actions that change customer data and to avoid creating duplicate tickets after a tool call times out.

Incorrect Answers

B:

The first proposal fits its assigned role because distinct interfaces simplify least-privilege grants and side-effect controls, while the second proposal serves the different purpose to support multiple simultaneous lookups without mixing their responses instead of the assigned second role because correlation IDs connect each observation to the exact invocation that requested it; their respective acceptance criteria are to require stricter approval for actions that change customer data and to avoid creating duplicate tickets after a tool call times out.

C:

The first proposal serves the different purpose to stop a directory lookup function from also gaining unrestricted database write access instead of the assigned first role because minimal capabilities reduce risk when generated arguments or retrieved text are hostile, while the second proposal fits its assigned role because stable operation identifiers let downstream systems detect repeated submissions; their respective acceptance criteria are to require stricter approval for actions that change customer data and to avoid creating duplicate tickets after a tool call times out.

D:

The first proposal serves the different purpose to support multiple simultaneous lookups without mixing their responses instead of the assigned first role because correlation IDs connect each observation to the exact invocation that requested it, while the second proposal serves the different purpose to stop a directory lookup function from also gaining unrestricted database write access instead of the assigned second role because minimal capabilities reduce risk when generated arguments or retrieved text are hostile; their respective acceptance criteria are to require stricter approval for actions that change customer data and to avoid creating duplicate tickets after a tool call times out.

Question 380

A document processing application is approaching release, but the team still needs to give Claude Code the project's build, test, and architecture conventions. What is the most appropriate implementation?

  1. Put repository guidance in CLAUDE.md
  2. Validate configuration at startup
  3. Version prompt templates alongside code
  4. Keep environment-specific settings external
  5. Pin a tested model identifier

Explanation

Correct Answer

A:

Repository memory files provide persistent instructional context to coding sessions; this behavior directly supports the need to give Claude Code the project's build, test, and architecture conventions. They are not an application secrets manager or the API's global system prompt.

Incorrect Answers

B:

For the alternate objective to fail fast if a model ID or tool endpoint is absent in production, startup checks surface missing required values before user traffic reaches a broken path; the feature here must give Claude Code the project's build, test, and architecture conventions, and a late runtime exception spreads one configuration defect across many requests.

C:

Source-controlled prompt assets make rollbacks, diffs, and evaluations possible; this applies to the different goal to trace a quality regression to the exact instruction revision used, not the requirement to give Claude Code the project's build, test, and architecture conventions; editing prompts in an unversioned dashboard hides causal history.

D:

When an application needs to deploy the same application build to staging and production with distinct endpoints, configuration injection separates environment differences from immutable application artifacts; the current capability must instead give Claude Code the project's build, test, and architecture conventions, while hard-coded URLs require risky code changes for promotion.

E:

Explicit model versions support reproducibility and controlled migrations; the supported outcome is to keep deployment behavior stable while evaluating new model releases separately, whereas this scenario requires the ability to give Claude Code the project's build, test, and architecture conventions; a floating alias can change behavior without an application code change.

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