Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 19

Question 180

During review of a product landscape research agent, an unmet criterion states that a long workflow preserves confirmed facts separately from speculation. Which measure resolves this specific gap?

  1. combine guesses and verified attributes in one unqualified summary
  2. pass a compact structured record of facts, decisions, pending tasks, and sources
  3. retrieve authoritative facts from a durable store when needed
  4. mark verified, user-reported, and unknown state explicitly

Explanation

Correct Answer

D:

For a product landscape research agent, epistemic labels stop a tentative suggestion becoming a confirmed action parameter; this directly establishes that a long workflow preserves confirmed facts separately from speculation instead of relying on an unrelated workflow control.

Incorrect Answers

A:

In a product landscape research agent, downstream operations may rely on an assumption as though it were checked; as a result, the stated requirement that a long workflow preserves confirmed facts separately from speculation would still not be met.

B:

Although focused handoffs reduce context size while protecting continuity-critical details, that only supports the distinct goal that a handoff retains only action-relevant context, not the requirement that a long workflow preserves confirmed facts separately from speculation in a product landscape research agent.

C:

Although re-reading source state restores precision that a condensed history cannot guarantee, that only supports the distinct goal that the agent can recover exact policy facts after compaction, not the requirement that a long workflow preserves confirmed facts separately from speculation in a product landscape research agent.

Question 181

During review of a contract clause extraction workflow, an unmet criterion states that an unresolved claim is qualified without inventing certainty. Which measure resolves this specific gap?

  1. label direct source facts separately from analytical conclusions
  2. report the disagreement and explain evidence quality or recency
  3. express the uncertainty and identify what verification is missing
  4. state a firm conclusion because the report needs a complete ending

Explanation

Correct Answer

C:

For a contract clause extraction workflow, calibrated language communicates a bounded evidence gap honestly; this directly establishes that an unresolved claim is qualified without inventing certainty instead of relying on an unrelated workflow control.

Incorrect Answers

A:

Although distinguishing evidence from interpretation prevents inference being misattributed to a document, that only supports the distinct goal that a report separates documented findings from its own inference, not the requirement that an unresolved claim is qualified without inventing certainty in a contract clause extraction workflow.

B:

Although explicit conflicts preserve uncertainty rather than manufacturing unanimity, that only supports the distinct goal that conflicting sources are not silently flattened into one answer, not the requirement that an unresolved claim is qualified without inventing certainty in a contract clause extraction workflow.

D:

In a contract clause extraction workflow, a polished narrative cannot supply missing corroboration; as a result, the stated requirement that an unresolved claim is qualified without inventing certainty would still not be met.

Question 182

The architect for a legacy service decomposition needs to ensure that a delegated skill should not fill the caller's active context. What should the team implement?

  1. configure the skill to run in an isolated forked context where supported
  2. create a documented reusable skill in SKILL.md
  3. define how the skill accepts and uses invocation arguments
  4. store project skills in the repository's .claude/skills structure
  5. load all investigation steps into the same conversation by default

Explanation

Correct Answer

A:

For a legacy service decomposition, a fork isolates exploratory work and returns a concise result to its caller; this directly establishes that a delegated skill should not fill the caller's active context instead of relying on an unrelated workflow control.

Incorrect Answers

B:

Although skills package instructions and supporting resources for invocation across similar tasks, that only supports the distinct goal that a repeatable code-review procedure is available as a named workflow, not the requirement that a delegated skill should not fill the caller's active context in a legacy service decomposition.

C:

Although parameterized workflows apply one reviewed procedure to different resources without hardcoding targets, that only supports the distinct goal that a shared skill can take a supplied target name or file path, not the requirement that a delegated skill should not fill the caller's active context in a legacy service decomposition.

D:

Although version-controlled skills evolve alongside tests and conventions with code review, that only supports the distinct goal that the repository can review and version its custom workflows, not the requirement that a delegated skill should not fill the caller's active context in a legacy service decomposition.

E:

In a legacy service decomposition, the caller's window accumulates unnecessary intermediate details; as a result, the stated requirement that a delegated skill should not fill the caller's active context would still not be met.

Question 183

In a policy document normalization service, the immediate design requirement is that a conclusion supported by model agreement is also tested against source evidence. Which change most directly establishes that behavior?

  1. calibrate confidence against known outcomes and source support
  2. require source and business-rule checks regardless of model agreement
  3. promote agreement alone to a guaranteed correctness signal
  4. put uncertain cases in an explicit review queue with priority and context
  5. route high-impact or low-evidence results to an authorized reviewer

Explanation

Correct Answer

B:

For a policy document normalization service, correlated model errors can survive consensus without external verification; this directly establishes that a conclusion supported by model agreement is also tested against source evidence instead of relying on an unrelated workflow control.

Incorrect Answers

A:

Although observed error rates and evidence quality are better than a fluent self-rating, that only supports the distinct goal that confidence scores reflect available evidence and uncertainty, not the requirement that a conclusion supported by model agreement is also tested against source evidence in a policy document normalization service.

C:

In a policy document normalization service, shared biases can produce identical wrong conclusions; as a result, the stated requirement that a conclusion supported by model agreement is also tested against source evidence would still not be met.

D:

Although triage preserves unattended processing for low-risk cases while catching exceptions, that only supports the distinct goal that an ambiguous item is separated from routine automation, not the requirement that a conclusion supported by model agreement is also tested against source evidence in a policy document normalization service.

E:

Although human approval supplies accountability where a mistaken model result has material cost, that only supports the distinct goal that a consequential claim receives independent human sign-off, not the requirement that a conclusion supported by model agreement is also tested against source evidence in a policy document normalization service.

Question 184

In a code policy conformance job, the acceptance criteria require that important workflow state survives context compression or process restarts and that a destructive tool receives human authorization at execution time. Which paired implementation meets both?

  1. persist authoritative task facts and artifact IDs outside ephemeral conversation text; and interrupt or deny the tool until an authorization callback approves it
  2. depend on the model remembering a prior value forever; and interrupt or deny the tool until an authorization callback approves it
  3. depend on the model remembering a prior value forever; and collect a general approval at the start of the conversation
  4. persist authoritative task facts and artifact IDs outside ephemeral conversation text; and collect a general approval at the start of the conversation

Explanation

Correct Answer

A:

In a code policy conformance job, external durable state can be reloaded and checked when the session window changes; alongside this, execution gates make the approval binding on the selected action and inputs, satisfying the distinct requirements that important workflow state survives context compression or process restarts and that a destructive tool receives human authorization at execution time.

Incorrect Answers

B:

In a code policy conformance job, execution gates make the approval binding on the selected action and inputs, but context limits and summarization can lose exact identifiers; therefore the separate requirement that important workflow state survives context compression or process restarts would remain unsatisfied.

C:

In a code policy conformance job, context limits and summarization can lose exact identifiers and blanket consent does not verify the later high-risk operation; the design would fail both conditions: important workflow state survives context compression or process restarts and a destructive tool receives human authorization at execution time.

D:

In a code policy conformance job, external durable state can be reloaded and checked when the session window changes, but blanket consent does not verify the later high-risk operation; therefore the separate requirement that a destructive tool receives human authorization at execution time would remain unsatisfied.

Question 185

An incident postmortem for a product landscape research agent requires controls so that a tool result includes enough fields to support audit and follow-on actions and that workload growth does not cause uncontrolled parallel fan-out. Which paired remedy covers both failure modes?

  1. return only a friendly summary with no source identifier; and limit concurrency and schedule independent jobs against execution budgets
  2. return only a friendly summary with no source identifier; and launch all possible child tasks at once without resource controls
  3. return structured identifiers, status, and provenance with the data; and limit concurrency and schedule independent jobs against execution budgets
  4. return structured identifiers, status, and provenance with the data; and launch all possible child tasks at once without resource controls

Explanation

Correct Answer

C:

In a product landscape research agent, machine-readable outputs let downstream stages validate exactly what was retrieved; alongside this, bounded orchestration balances throughput with tool quotas and downstream service capacity, satisfying the distinct requirements that a tool result includes enough fields to support audit and follow-on actions and that workload growth does not cause uncontrolled parallel fan-out.

Incorrect Answers

A:

In a product landscape research agent, bounded orchestration balances throughput with tool quotas and downstream service capacity, but a paraphrase hides the record needed for verification or follow-up; therefore the separate requirement that a tool result includes enough fields to support audit and follow-on actions would remain unsatisfied.

B:

In a product landscape research agent, a paraphrase hides the record needed for verification or follow-up and unbounded fan-out can overwhelm tools and blow through cost budgets; the design would fail both conditions: a tool result includes enough fields to support audit and follow-on actions and workload growth does not cause uncontrolled parallel fan-out.

D:

In a product landscape research agent, machine-readable outputs let downstream stages validate exactly what was retrieved, but unbounded fan-out can overwhelm tools and blow through cost budgets; therefore the separate requirement that workload growth does not cause uncontrolled parallel fan-out would remain unsatisfied.

Question 186

In an order-status triage agent, the acceptance criteria require that a broad tool catalog does not overwhelm a small task and that an order adjustment uses the current order record. Which paired implementation meets both?

  1. load all available MCP tools into every specialist turn; and issue the update using details inferred only from conversation
  2. supply only context-relevant tools for the current agent; and issue the update using details inferred only from conversation
  3. load all available MCP tools into every specialist turn; and require the order lookup result before calculating or applying a change
  4. supply only context-relevant tools for the current agent; and require the order lookup result before calculating or applying a change
  5. implement only supply only context-relevant tools for the current agent; make no corresponding change for the second requirement

Explanation

Correct Answer

D:

In an order-status triage agent, reducing irrelevant tool definitions improves selection and saves context; alongside this, the dependent operation needs authoritative state and must not guess from the customer's description, satisfying the distinct requirements that a broad tool catalog does not overwhelm a small task and that an order adjustment uses the current order record.

Incorrect Answers

A:

In an order-status triage agent, tool bloat makes the available actions harder to discriminate and unverified user text can reference the wrong order or outdated state; the design would fail both conditions: a broad tool catalog does not overwhelm a small task and an order adjustment uses the current order record.

B:

In an order-status triage agent, reducing irrelevant tool definitions improves selection and saves context, but unverified user text can reference the wrong order or outdated state; therefore the separate requirement that an order adjustment uses the current order record would remain unsatisfied.

C:

In an order-status triage agent, the dependent operation needs authoritative state and must not guess from the customer's description, but tool bloat makes the available actions harder to discriminate; therefore the separate requirement that a broad tool catalog does not overwhelm a small task would remain unsatisfied.

E:

In an order-status triage agent, reducing irrelevant tool definitions improves selection and saves context, but the independent mechanism for the other criterion is absent: the dependent operation needs authoritative state and must not guess from the customer's description; as written, the design does not ensure that an order adjustment uses the current order record.

Question 187

Production readiness for an expense receipt capture system depends on ensuring that a truncated response is not mistaken for successful completion and that a failed extraction is retried with specific corrective information. Select the most appropriate combined approach.

  1. interpret every non-tool stop reason as a successful final answer; and repeat the same vague prompt after each failure
  2. handle max_tokens as truncation rather than end_turn; and feed field-level validation errors back in a bounded retry
  3. interpret every non-tool stop reason as a successful final answer; and feed field-level validation errors back in a bounded retry
  4. handle max_tokens as truncation rather than end_turn; and repeat the same vague prompt after each failure

Explanation

Correct Answer

B:

In an expense receipt capture system, reaching an output limit is a stop condition but does not establish task completion; alongside this, targeted feedback identifies what the next attempt must correct, satisfying the distinct requirements that a truncated response is not mistaken for successful completion and that a failed extraction is retried with specific corrective information.

Incorrect Answers

A:

In an expense receipt capture system, a length-limited result may be incomplete or unusable and unchanged requests do not reveal which constraint was violated; the design would fail both conditions: a truncated response is not mistaken for successful completion and a failed extraction is retried with specific corrective information.

C:

In an expense receipt capture system, targeted feedback identifies what the next attempt must correct, but a length-limited result may be incomplete or unusable; therefore the separate requirement that a truncated response is not mistaken for successful completion would remain unsatisfied.

D:

In an expense receipt capture system, reaching an output limit is a stop condition but does not establish task completion, but unchanged requests do not reveal which constraint was violated; therefore the separate requirement that a failed extraction is retried with specific corrective information would remain unsatisfied.

Question 188

In a monorepo testing overhaul, the immediate design requirement is that path-scoped rules are loaded for all intended source directories. Which change most directly establishes that behavior?

  1. add more prose telling the model to remember the missing rule
  2. create separate path-specific rule files for each package
  3. keep project-wide security guidance in shared instructions rather than a narrow path glob
  4. update the rule's paths glob to include the actual target paths

Explanation

Correct Answer

D:

For a monorepo testing overhaul, correct glob coverage determines whether conditional guidance is in context; this directly establishes that path-scoped rules are loaded for all intended source directories instead of relying on an unrelated workflow control.

Incorrect Answers

A:

In a monorepo testing overhaul, instructions outside the matched scope still will not be loaded; as a result, the stated requirement that path-scoped rules are loaded for all intended source directories would still not be met.

B:

Although scoped conventions avoid contradictory patterns between codebases sharing a monorepo, that only supports the distinct goal that two nested packages follow distinct framework conventions, not the requirement that path-scoped rules are loaded for all intended source directories in a monorepo testing overhaul.

C:

Although cross-cutting obligations should not disappear when an agent changes directories, that only supports the distinct goal that a common security rule remains active across the project, not the requirement that path-scoped rules are loaded for all intended source directories in a monorepo testing overhaul.

Question 189

An incident postmortem for a quarterly market-intelligence coordinator requires controls so that a conversational task does not invoke tools needlessly and that a truncated response is not mistaken for successful completion. Which paired remedy covers both failure modes?

  1. use automatic tool choice when a call is optional; and handle max_tokens as truncation rather than end_turn
  2. force a tool on every message regardless of intent; and handle max_tokens as truncation rather than end_turn
  3. use automatic tool choice when a call is optional; and interpret every non-tool stop reason as a successful final answer
  4. force a tool on every message regardless of intent; and interpret every non-tool stop reason as a successful final answer

Explanation

Correct Answer

A:

In a quarterly market-intelligence coordinator, auto allows the model to answer directly when no external side effect or fresh data is needed; alongside this, reaching an output limit is a stop condition but does not establish task completion, satisfying the distinct requirements that a conversational task does not invoke tools needlessly and that a truncated response is not mistaken for successful completion.

Incorrect Answers

B:

In a quarterly market-intelligence coordinator, reaching an output limit is a stop condition but does not establish task completion, but mandatory calls add latency and can invoke irrelevant operations; therefore the separate requirement that a conversational task does not invoke tools needlessly would remain unsatisfied.

C:

In a quarterly market-intelligence coordinator, auto allows the model to answer directly when no external side effect or fresh data is needed, but a length-limited result may be incomplete or unusable; therefore the separate requirement that a truncated response is not mistaken for successful completion would remain unsatisfied.

D:

In a quarterly market-intelligence coordinator, mandatory calls add latency and can invoke irrelevant operations and a length-limited result may be incomplete or unusable; the design would fail both conditions: a conversational task does not invoke tools needlessly and a truncated response is not mistaken for successful completion.

Question 190

The release gate for an order-cancellation service requires that a selector accepts only supported operation names; it also requires that a truncated response is not mistaken for successful completion. Which TWO actions meet the specified checks? Choose TWO.

  1. declare an operation enum in the schema and reject values outside it in the tool handler
  2. handle max_tokens as truncation rather than end_turn
  3. interpret every non-tool stop reason as a successful final answer
  4. ask the model to invent a suitable action string

Explanation

Correct Answers

A:

For an order-cancellation service, schema constraints guide valid selections and runtime validation blocks unsupported operation values, which directly establishes that a selector accepts only supported operation names; the separate requirement that a truncated response is not mistaken for successful completion is handled by a different action.

B:

For an order-cancellation service, reaching an output limit is a stop condition but does not establish task completion, which directly establishes that a truncated response is not mistaken for successful completion; the separate requirement that a selector accepts only supported operation names is handled by a different action.

Incorrect Answers

C:

For an order-cancellation service, a length-limited result may be incomplete or unusable, so this option cannot establish that a truncated response is not mistaken for successful completion; correctly addressing a selector accepts only supported operation names elsewhere does not repair that deficiency.

D:

For an order-cancellation service, free-form labels can name unimplemented operations, and a prompt is not an enforced allowlist, so this option cannot establish that a selector accepts only supported operation names; correctly addressing a truncated response is not mistaken for successful completion elsewhere does not repair that deficiency.

Claude Certified Architect - Foundations (CCAR-F) (CCA-F) Practice Tests