Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 40

Question 390

A production-readiness check for an incident response component requires the service to preserve evidence and data lineage when one agent delegates a subtask to another. Which approach passes this check?

  1. Pass structured handoff artifacts between workers
  2. Use an explicit completion condition and step budget
  3. Use a deterministic workflow
  4. Introduce a human approval checkpoint

Explanation

Correct Answer

A:

Defined input and output schemas make it possible to attribute conclusions and detect missing fields at the handoff; this behavior directly supports the need to preserve evidence and data lineage when one agent delegates a subtask to another. Free-form chat summaries can omit identifiers required by the next worker.

Incorrect Answers

B:

For the alternate objective 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 feature here must preserve evidence and data lineage when one agent delegates a subtask to another, and merely asking the model to finish quickly does not enforce execution limits.

C:

A fixed pipeline keeps control flow in application code, making execution order reproducible and easy to audit; this applies to the different goal to execute an unchanged sequence of extraction, validation, and publishing steps, not the requirement to preserve evidence and data lineage when one agent delegates a subtask to another; a free-running agent adds planning variability without helping when every step is already known.

D:

When an application needs to prevent an autonomous investigation from issuing irreversible account changes without review, a human gate separates analysis from externally consequential action and gives a reviewer a veto; the current capability must instead preserve evidence and data lineage when one agent delegates a subtask to another, while a system-prompt request to be careful is not an enforceable approval workflow.

Question 391

The owner of a Claude Code repository integration wants to continue an interrupted refactor while retaining relevant prior decisions. Which proposal provides the required mechanism?

  1. Connect an MCP server to Claude Code
  2. Use headless invocation for automation
  3. Create a reusable Skill
  4. Review and resume a saved session

Explanation

Correct Answer

D:

Session continuation can reestablish conversational state without restarting the task from scratch; this behavior directly supports the need to continue an interrupted refactor while retaining relevant prior decisions. The developer still needs to recheck whether repository files changed in the meantime.

Incorrect Answers

A:

When an application needs to expose approved project-specific actions without copying credentials into prompts, an MCP integration declares external capabilities and their transport while access remains controlled by the server; the current capability must instead continue an interrupted refactor while retaining relevant prior decisions, while a server connection does not automatically make every operation safe or authorized.

B:

Noninteractive execution lets scripts integrate agent work into existing pipelines with controlled formats; the supported outcome is to run Claude Code inside a CI job and capture its output without interactive prompts, whereas this scenario requires the ability to continue an interrupted refactor while retaining relevant prior decisions; treating interactive terminal output as a stable machine interface is fragile.

C:

This approach serves the separate need to make a specialised review procedure available as task-specific instructions and assets because a Skill packages a repeatable capability that Claude can load when relevant to the work; it does not establish the outcome to continue an interrupted refactor while retaining relevant prior decisions, and a one-off chat turn is not a maintainable shared workflow definition.

Question 392

An engineering productivity team is evaluating two disjoint use cases. For one use case, it must enforce approval, execution, and confirmation order in a compliance workflow; for the other, it must continue a support investigation after many rounds of log inspection. Which paired choices are justified?

  1. First component: Use a plan-execute-review loop. Second component: Summarize long-running state with checkpoints.
  2. First component: Use an independent verifier step. Second component: Use a plan-execute-review loop.
  3. First component: Prefer a finite state machine for regulated steps. Second component: Summarize long-running state with checkpoints.
  4. First component: Prefer a finite state machine for regulated steps. Second component: Use an independent verifier step.

Explanation

Correct Answer

C:

The first proposal fits its assigned role because explicit states and transitions control who may advance a case independently of generated language, while the second proposal fits its assigned role because periodic compaction preserves decisions, open questions, and evidence references within the context budget; their respective acceptance criteria are to enforce approval, execution, and confirmation order in a compliance workflow and to continue a support investigation after many rounds of log inspection.

Incorrect Answers

A:

The first proposal serves the different purpose to manage a complex migration that requires checkpoints after several tool actions instead of the assigned first role because separating plan, execution, and review makes deviations visible before a final outcome is accepted, while the second proposal fits its assigned role because periodic compaction preserves decisions, open questions, and evidence references within the context budget; their respective acceptance criteria are to enforce approval, execution, and confirmation order in a compliance workflow and to continue a support investigation after many rounds of log inspection.

B:

The first proposal serves the different purpose to detect mistakes in an agent-generated SQL migration before running it instead of the assigned first role because a separate validation stage checks the artifact against explicit tests and constraints, limiting self-confirming errors, while the second proposal serves the different purpose to manage a complex migration that requires checkpoints after several tool actions instead of the assigned second role because separating plan, execution, and review makes deviations visible before a final outcome is accepted; their respective acceptance criteria are to enforce approval, execution, and confirmation order in a compliance workflow and to continue a support investigation after many rounds of log inspection.

D:

The first proposal fits its assigned role because explicit states and transitions control who may advance a case independently of generated language, while the second proposal serves the different purpose to detect mistakes in an agent-generated SQL migration before running it instead of the assigned second role because a separate validation stage checks the artifact against explicit tests and constraints, limiting self-confirming errors; their respective acceptance criteria are to enforce approval, execution, and confirmation order in a compliance workflow and to continue a support investigation after many rounds of log inspection.

Question 393

A design meeting for a model platform platform concludes that two isolated services require different behavior: one must avoid treating a tool call or thinking block as user-visible answer text, and another must estimate model-token input demand before launching a large processing job. Which arrangement meets the specification?

  1. First component: Parse responses by content block type. Second component: Use a queue for burst absorption.
  2. First component: Parse responses by content block type. Second component: Use the Messages count_tokens capability.
  3. First component: Choose SSE for one-way streamed responses. Second component: Use the Messages count_tokens capability.
  4. First component: Use a queue for burst absorption. Second component: Choose SSE for one-way streamed responses.

Explanation

Correct Answer

B:

The first proposal fits its assigned role because structured outputs can contain different block kinds that need type-aware handling, while the second proposal fits its assigned role because token counting helps check input budgets before generation and cost planning; their respective acceptance criteria are to avoid treating a tool call or thinking block as user-visible answer text and to estimate model-token input demand before launching a large processing job.

Incorrect Answers

A:

The first proposal fits its assigned role because structured outputs can contain different block kinds that need type-aware handling, while the second proposal serves the different purpose to smooth sudden ingestion spikes before workers call a capacity-limited model API instead of the assigned second role because queueing decouples producers from consumer throughput and enables controlled retries; their respective acceptance criteria are to avoid treating a tool call or thinking block as user-visible answer text and to estimate model-token input demand before launching a large processing job.

C:

The first proposal serves the different purpose to deliver tokens to a browser without maintaining a bidirectional command channel instead of the assigned first role because server-sent events provide ordered server-to-client updates over an HTTP stream, while the second proposal fits its assigned role because token counting helps check input budgets before generation and cost planning; their respective acceptance criteria are to avoid treating a tool call or thinking block as user-visible answer text and to estimate model-token input demand before launching a large processing job.

D:

The first proposal serves the different purpose to smooth sudden ingestion spikes before workers call a capacity-limited model API instead of the assigned first role because queueing decouples producers from consumer throughput and enables controlled retries, while the second proposal serves the different purpose to deliver tokens to a browser without maintaining a bidirectional command channel instead of the assigned second role because server-sent events provide ordered server-to-client updates over an HTTP stream; their respective acceptance criteria are to avoid treating a tool call or thinking block as user-visible answer text and to estimate model-token input demand before launching a large processing job.

Question 394

An architecture review of a production API product lists two distinct outcomes: the first subsystem must avoid exposing raw production database credentials directly to an agent, and a second subsystem must avoid burying the relevant contract clause beneath hundreds of low-value passages. Which pair delivers these outcomes?

  1. First component: Isolate tools behind a service layer. Second component: Use bounded retrieval results.
  2. First component: Treat retrieval results as untrusted data. Second component: Build a typed output envelope.
  3. First component: Build a typed output envelope. Second component: Use bounded retrieval results.
  4. First component: Isolate tools behind a service layer. Second component: Treat retrieval results as untrusted data.

Explanation

Correct Answer

A:

The first proposal fits its assigned role because a mediation service validates identity, arguments, authorization, and audit logs for each requested action, while the second proposal fits its assigned role because ranking and a context budget prioritize high-quality evidence while controlling input tokens; their respective acceptance criteria are to avoid exposing raw production database credentials directly to an agent and to avoid burying the relevant contract clause beneath hundreds of low-value passages.

Incorrect Answers

B:

The first proposal serves the different purpose to prevent a malicious help article from ordering the model to reveal secrets instead of the assigned first role because retrieved content should be framed as evidence, never upgraded to new tool or system authority, while the second proposal serves the different purpose to send extracted contract fields into a database that rejects unexpected shapes instead of the assigned second role because an explicit schema with required fields, types, and allowed values makes integration errors detectable; their respective acceptance criteria are to avoid exposing raw production database credentials directly to an agent and to avoid burying the relevant contract clause beneath hundreds of low-value passages.

C:

The first proposal serves the different purpose to send extracted contract fields into a database that rejects unexpected shapes instead of the assigned first role because an explicit schema with required fields, types, and allowed values makes integration errors detectable, while the second proposal fits its assigned role because ranking and a context budget prioritize high-quality evidence while controlling input tokens; their respective acceptance criteria are to avoid exposing raw production database credentials directly to an agent and to avoid burying the relevant contract clause beneath hundreds of low-value passages.

D:

The first proposal fits its assigned role because a mediation service validates identity, arguments, authorization, and audit logs for each requested action, while the second proposal serves the different purpose to prevent a malicious help article from ordering the model to reveal secrets instead of the assigned second role because retrieved content should be framed as evidence, never upgraded to new tool or system authority; their respective acceptance criteria are to avoid exposing raw production database credentials directly to an agent and to avoid burying the relevant contract clause beneath hundreds of low-value passages.

Question 395

A developer is reviewing alternatives for an application security feature that must attribute tool actions to the right workload rather than a developer's personal account. Which option is a defensible fit?

  1. Separate human and service identities
  2. Protect secrets in error traces
  3. Load keys from a managed secret service
  4. Rotate a compromised API key promptly

Explanation

Correct Answer

A:

Distinct principals simplify least privilege, revocation, and audit attribution; this behavior directly supports the need to attribute tool actions to the right workload rather than a developer's personal account. A shared personal login makes operational accountability difficult.

Incorrect Answers

B:

Redaction at logging boundaries avoids accidentally copying credentials to broad observability systems; despite being useful when a service must prevent stack traces and debugging telemetry from disclosing tokens, it is not a solution to attribute tool actions to the right workload rather than a developer's personal account; encrypted storage does not help when secrets are printed into logs.

C:

For the alternate objective to avoid committing provider credentials inside a shared repository, secrets managers support controlled retrieval, rotation, and access logs; the feature here must attribute tool actions to the right workload rather than a developer's personal account, and obscuring a literal string in source is not adequate key protection.

D:

Revocation plus replacement ends use of the leaked credential and requires checking for misuse; this applies to the different goal to cut off an exposed token discovered in CI logs, not the requirement to attribute tool actions to the right workload rather than a developer's personal account; merely deleting a log message leaves the old token valid.

Question 396

A technical lead requests a solution that can enable a reviewer to reconstruct evidence after context has been compacted within a conversation design project. Which design should be chosen?

  1. Compact completed task history
  2. Anchor summaries to source identifiers
  3. Separate long-term memory from active scratchpad
  4. Prioritize high-signal retrieved passages
  5. Preserve instruction and evidence boundaries

Explanation

Correct Answer

B:

Stable document IDs and passage references preserve provenance beyond the shortened working memory; this behavior directly supports the need to enable a reviewer to reconstruct evidence after context has been compacted. A bare compressed claim cannot show which source established it.

Incorrect Answers

A:

A structured summary preserves established facts, pending work, and unresolved assumptions in fewer tokens; the supported outcome is to let an assistant continue a long incident investigation without losing decisions, whereas this scenario requires the ability to enable a reviewer to reconstruct evidence after context has been compacted; raw truncation may remove the reason an earlier decision was made.

C:

Explicit storage and freshness rules govern what is retained and when it is reintroduced; despite being useful when a service must prevent outdated user preferences from being treated as current task instructions, it is not a solution to enable a reviewer to reconstruct evidence after context has been compacted; an unbounded persistent transcript can accumulate contradictory assumptions.

D:

For the alternate objective to fit a large knowledge base into the context of a customer answer, ranking by relevance and authority places the most useful evidence ahead of incidental matches; the feature here must enable a reviewer to reconstruct evidence after context has been compacted, and dumping every document into the prompt can obscure the actual answer.

E:

Context assembly must label external text as data and avoid granting it authority over tool use; this applies to the different goal to keep an embedded webpage instruction from becoming higher-priority application policy, not the requirement to enable a reviewer to reconstruct evidence after context has been compacted; quotation formatting alone does not enforce sandboxing or authorization.

Question 397

A production-readiness check for a prompt engineering component requires the service to distinguish near-duplicate support intents that a broad rule routinely confuses. Which approach passes this check?

  1. State one measurable outcome per instruction
  2. Specify output format in the prompt
  3. Use few-shot boundary cases
  4. Constrain response audience and purpose

Explanation

Correct Answer

C:

Examples illustrating each side of a decision boundary communicate nuanced classification criteria; this behavior directly supports the need to distinguish near-duplicate support intents that a broad rule routinely confuses. Many unrelated demonstrations can make the rule harder to infer.

Incorrect Answers

A:

Clear task goals and acceptance criteria reduce competing interpretations; the supported outcome is to stop a summarizer from producing unrelated analysis instead of required action items, whereas this scenario requires the ability to distinguish near-duplicate support intents that a broad rule routinely confuses; a vague request for a good summary provides no testable target.

B:

This approach serves the separate need to obtain named fields in a consistent response layout for subsequent processing because explicit field descriptions guide generation toward the consumer's expected structure; it does not establish the outcome to distinguish near-duplicate support intents that a broad rule routinely confuses, and formatting guidance still requires machine validation at the boundary.

D:

For the alternate objective to generate a technical incident summary suitable for an on-call engineer instead of an executive, audience constraints determine useful terminology, specificity, and level of detail; the feature here must distinguish near-duplicate support intents that a broad rule routinely confuses, and style guidance alone cannot make inaccurate facts true.

Question 398

A technical lead requests a solution that can perform open-ended troubleshooting where tool selection changes after each observation within a tool integration project. Which design should be chosen?

  1. Use an autonomous agent for exploratory work
  2. Publish an MCP server for reusable integrations
  3. Choose a built-in provider tool
  4. Use system or repository instructions for stable behavior

Explanation

Correct Answer

A:

Model-directed planning supports adaptive sequences when the next action is not known in advance; this behavior directly supports the need to perform open-ended troubleshooting where tool selection changes after each observation. A rigid workflow struggles with unforeseen branches without extensive coding.

Incorrect Answers

B:

MCP standardizes tool and resource discovery across compatible applications; this applies to the different goal to make one internal ticketing connector available to several agent clients, not the requirement to perform open-ended troubleshooting where tool selection changes after each observation; duplicated per-client wrappers increase maintenance across a growing ecosystem.

C:

When an application needs to use an already supported capability without hosting a duplicate custom service, provider-managed execution can reduce implementation work when its scope matches the requirement; the current capability must instead perform open-ended troubleshooting where tool selection changes after each observation, while built-in tools may not expose a proprietary enterprise data source.

D:

Instructions shape model behavior and task framing without executing external operations; the supported outcome is to make a coding or API assistant follow organization-wide conventions, whereas this scenario requires the ability to perform open-ended troubleshooting where tool selection changes after each observation; descriptive rules do not replace programmatic validation or IAM.

Question 399

A production-readiness check for an incident response component requires the service to analyze several independent document sections concurrently and combine the results. Which approach passes this check?

  1. Introduce a human approval checkpoint
  2. Use a deterministic workflow
  3. Use an agentic decision loop
  4. Use parallel fan-out and a reconciliation step

Explanation

Correct Answer

D:

Parallel tasks reduce critical-path latency when they do not depend on one another; a join step resolves conflicting findings; this behavior directly supports the need to analyze several independent document sections concurrently and combine the results. A serial chain makes independent stages wait without improving correctness.

Incorrect Answers

A:

For the alternate objective to prevent an autonomous investigation from issuing irreversible account changes without review, a human gate separates analysis from externally consequential action and gives a reviewer a veto; the feature here must analyze several independent document sections concurrently and combine the results, and a system-prompt request to be careful is not an enforceable approval workflow.

B:

A fixed pipeline keeps control flow in application code, making execution order reproducible and easy to audit; this applies to the different goal to execute an unchanged sequence of extraction, validation, and publishing steps, not the requirement to analyze several independent document sections concurrently and combine the results; a free-running agent adds planning variability without helping when every step is already known.

C:

When an application needs to investigate an open-ended incident where the next query depends on earlier findings, a model-directed loop can choose subsequent actions based on tool observations until it reaches a stopping condition; the current capability must instead analyze several independent document sections concurrently and combine the results, while a predetermined chain cannot reliably anticipate all branches of an exploratory diagnosis.

Question 400

A technical lead requests a solution that can cut off an exposed token discovered in CI logs within a credential management project. Which design should be chosen?

  1. Rotate a compromised API key promptly
  2. Use short-lived scoped credentials where supported
  3. Load keys from a managed secret service
  4. Authorize by requested resource and action

Explanation

Correct Answer

A:

Revocation plus replacement ends use of the leaked credential and requires checking for misuse; this behavior directly supports the need to cut off an exposed token discovered in CI logs. Merely deleting a log message leaves the old token valid.

Incorrect Answers

B:

Expiring identity grants reduce the time and permissions available to an attacker; the supported outcome is to limit the impact of a temporary worker credential being stolen, whereas this scenario requires the ability to cut off an exposed token discovered in CI logs; permanent shared administrator keys expand blast radius.

C:

This approach serves the separate need to avoid committing provider credentials inside a shared repository because secrets managers support controlled retrieval, rotation, and access logs; it does not establish the outcome to cut off an exposed token discovered in CI logs, and obscuring a literal string in source is not adequate key protection.

D:

Authorization checks must evaluate scope for each operation, not only whether a login exists; despite being useful when a service must deny access to a different customer's records even when the caller is authenticated, it is not a solution to cut off an exposed token discovered in CI logs; successful authentication is not equivalent to unrestricted data access.

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