Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 56

Question 550

An incident postmortem for an expense receipt capture system requires controls so that tool output is normalized before it enters the agent context and that a broad tool catalog does not overwhelm a small task. Which paired remedy covers both failure modes?

  1. preprocess only the original user message; and load all available MCP tools into every specialist turn
  2. use a PostToolUse processing hook to standardize the returned data; and supply only context-relevant tools for the current agent
  3. use a PostToolUse processing hook to standardize the returned data; and load all available MCP tools into every specialist turn
  4. preprocess only the original user message; and supply only context-relevant tools for the current agent

Explanation

Correct Answer

B:

In an expense receipt capture system, post-execution interception can clean result fields and attach useful status without changing the model's tool selection; alongside this, reducing irrelevant tool definitions improves selection and saves context, satisfying the distinct requirements that tool output is normalized before it enters the agent context and that a broad tool catalog does not overwhelm a small task.

Incorrect Answers

A:

In an expense receipt capture system, input normalization does not correct inconsistent tool outputs and tool bloat makes the available actions harder to discriminate; the design would fail both conditions: tool output is normalized before it enters the agent context and a broad tool catalog does not overwhelm a small task.

C:

In an expense receipt capture system, post-execution interception can clean result fields and attach useful status without changing the model's tool selection, 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.

D:

In an expense receipt capture system, reducing irrelevant tool definitions improves selection and saves context, but input normalization does not correct inconsistent tool outputs; therefore the separate requirement that tool output is normalized before it enters the agent context would remain unsatisfied.

Question 551

In a technical review of a code-reference retrieval assistant, the architect must address the following constraints together: specialists return comparable artifacts for aggregation; separate code modules can be inspected in parallel. Which change set is most defensible?

  1. merge narrative responses by concatenating their text; and spawn independent specialists in the same delegation stage
  2. give subagents a common results contract including claim, evidence, and status; and spawn independent specialists in the same delegation stage
  3. merge narrative responses by concatenating their text; and make the second specialist wait for the first even though their modules are independent
  4. give subagents a common results contract including claim, evidence, and status; and make the second specialist wait for the first even though their modules are independent

Explanation

Correct Answer

B:

In a code-reference retrieval assistant, a typed handoff lets the coordinator compare like fields and detect missing support; alongside this, concurrent subagents can examine non-overlapping modules before the coordinator merges their reports, satisfying the distinct requirements that specialists return comparable artifacts for aggregation and that separate code modules can be inspected in parallel.

Incorrect Answers

A:

In a code-reference retrieval assistant, concurrent subagents can examine non-overlapping modules before the coordinator merges their reports, but unstructured concatenation hides provenance and duplicate or conflicting claims; therefore the separate requirement that specialists return comparable artifacts for aggregation would remain unsatisfied.

C:

In a code-reference retrieval assistant, unstructured concatenation hides provenance and duplicate or conflicting claims and serialization wastes time without supplying a dependency; the design would fail both conditions: specialists return comparable artifacts for aggregation and separate code modules can be inspected in parallel.

D:

In a code-reference retrieval assistant, a typed handoff lets the coordinator compare like fields and detect missing support, but serialization wastes time without supplying a dependency; therefore the separate requirement that separate code modules can be inspected in parallel would remain unsatisfied.

Question 552

An audit of a quarterly market-intelligence coordinator identifies two unmet requirements: file writes are blocked outside an approved workspace and agents exchange outputs without losing important structure. Select TWO actions that remedy those particular deficiencies.

  1. tell the model not to leave the workspace
  2. check resolved target paths in the pre-tool policy gate
  3. pass raw chat logs between every stage
  4. define artifact schemas and handoff contracts between roles

Explanation

Correct Answers

B:

For a quarterly market-intelligence coordinator, a boundary check on the actual tool argument enforces the permitted filesystem scope, which directly establishes that file writes are blocked outside an approved workspace; the separate requirement that agents exchange outputs without losing important structure is handled by a different action.

D:

For a quarterly market-intelligence coordinator, bounded interfaces make missing fields and incompatible assumptions visible, which directly establishes that agents exchange outputs without losing important structure; the separate requirement that file writes are blocked outside an approved workspace is handled by a different action.

Incorrect Answers

A:

For a quarterly market-intelligence coordinator, language instructions do not intercept a disallowed filesystem operation, so this option cannot establish that file writes are blocked outside an approved workspace; correctly addressing agents exchange outputs without losing important structure elsewhere does not repair that deficiency.

C:

For a quarterly market-intelligence coordinator, unstructured context is noisy and can obscure the decision fields, so this option cannot establish that agents exchange outputs without losing important structure; correctly addressing file writes are blocked outside an approved workspace elsewhere does not repair that deficiency.

Question 553

The release checklist for a shipping-exception helpdesk explicitly checks that a status field can contain only approved business states; it also checks that an uncertain high-impact decision is routed for review instead of automatic execution. What coordinated solution meets the checklist?

  1. enforce the approved status enum with structured output and validate it in the consumer; and proceed whenever the assistant's prose sounds certain
  2. accept any model-generated status description; and set a clear evidence or confidence boundary and require review below it
  3. accept any model-generated status description; and proceed whenever the assistant's prose sounds certain
  4. enforce the approved status enum with structured output and validate it in the consumer; and set a clear evidence or confidence boundary and require review below it

Explanation

Correct Answer

D:

In a shipping-exception helpdesk, a constrained schema expresses the allowed labels and downstream validation rejects out-of-contract values; alongside this, risk-weighted gates limit autonomous actions when support is inadequate, satisfying the distinct requirements that a status field can contain only approved business states and that an uncertain high-impact decision is routed for review instead of automatic execution.

Incorrect Answers

A:

In a shipping-exception helpdesk, a constrained schema expresses the allowed labels and downstream validation rejects out-of-contract values, but fluent language is not a calibrated reliability measure; therefore the separate requirement that an uncertain high-impact decision is routed for review instead of automatic execution would remain unsatisfied.

B:

In a shipping-exception helpdesk, risk-weighted gates limit autonomous actions when support is inadequate, but free-form status phrases can violate a fixed business-state contract; therefore the separate requirement that a status field can contain only approved business states would remain unsatisfied.

C:

In a shipping-exception helpdesk, free-form status phrases can violate a fixed business-state contract and fluent language is not a calibrated reliability measure; the design would fail both conditions: a status field can contain only approved business states and an uncertain high-impact decision is routed for review instead of automatic execution.

Question 554

In an order-cancellation service, the engineering team must guarantee that source identity survives multi-agent research handoffs. What is the most appropriate control?

  1. attach source IDs and relevant excerpts to individual claims
  2. carry citation identifiers and retrieval context through synthesis
  3. report the disagreement and explain evidence quality or recency
  4. label direct source facts separately from analytical conclusions
  5. strip all citation fields before sending the writer its inputs

Explanation

Correct Answer

B:

For an order-cancellation service, provenance metadata must remain attached when intermediate summaries are merged; this directly establishes that source identity survives multi-agent research handoffs instead of relying on an unrelated workflow control.

Incorrect Answers

A:

Although claim-level provenance enables an auditor to verify each assertion, that only supports the distinct goal that a synthesized statement can be traced to its supporting source, not the requirement that source identity survives multi-agent research handoffs in an order-cancellation service.

C:

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 source identity survives multi-agent research handoffs in an order-cancellation service.

D:

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 source identity survives multi-agent research handoffs in an order-cancellation service.

E:

In an order-cancellation service, the writer cannot reliably reconstruct which source supports each conclusion; as a result, the stated requirement that source identity survives multi-agent research handoffs would still not be met.

Question 555

For a pull-request regression detection job, acceptance depends on demonstrating that an automation expects a defined JSON result schema. Which approach is the strongest fit?

  1. ask for valid JSON in prose but accept any text result
  2. persist structured errors, relevant logs, and a bounded diagnostic summary
  3. use the CLI structured-output schema option for the response
  4. scope runner credentials, worktree access, and allowed tools to its task

Explanation

Correct Answer

C:

For a pull-request regression detection job, schema-constrained output is more robust for downstream integrations than prompt-only formatting; this directly establishes that an automation expects a defined JSON result schema instead of relying on an unrelated workflow control.

Incorrect Answers

A:

In a pull-request regression detection job, a natural-language instruction does not guarantee parseable or complete fields; as a result, the stated requirement that an automation expects a defined JSON result schema would still not be met.

B:

Although stable artifacts help maintainers reproduce the problem without rerunning a chat, that only supports the distinct goal that CI failures provide actionable evidence to maintainers, not the requirement that an automation expects a defined JSON result schema in a pull-request regression detection job.

D:

Although CI identity and tool permissions control real write capabilities independent of model instructions, that only supports the distinct goal that a pull-request bot runs with minimum necessary authority, not the requirement that an automation expects a defined JSON result schema in a pull-request regression detection job.

Question 556

Tests for a financial statement data extractor will pass only if every high-impact tool invocation produces an audit event and if a conversational task does not invoke tools needlessly. Which proposed design directly satisfies both checks?

  1. record the tool name, permitted inputs, outcome, and task identifier in a hook; and use automatic tool choice when a call is optional
  2. infer tool activity from the final natural-language answer; and use automatic tool choice when a call is optional
  3. infer tool activity from the final natural-language answer; and force a tool on every message regardless of intent
  4. record the tool name, permitted inputs, outcome, and task identifier in a hook; and force a tool on every message regardless of intent

Explanation

Correct Answer

A:

In a financial statement data extractor, runtime instrumentation observes actual proposed or completed tool activity; alongside this, auto allows the model to answer directly when no external side effect or fresh data is needed, satisfying the distinct requirements that every high-impact tool invocation produces an audit event and that a conversational task does not invoke tools needlessly.

Incorrect Answers

B:

In a financial statement data extractor, auto allows the model to answer directly when no external side effect or fresh data is needed, but a summary can omit failed calls or delegated actions; therefore the separate requirement that every high-impact tool invocation produces an audit event would remain unsatisfied.

C:

In a financial statement data extractor, a summary can omit failed calls or delegated actions and mandatory calls add latency and can invoke irrelevant operations; the design would fail both conditions: every high-impact tool invocation produces an audit event and a conversational task does not invoke tools needlessly.

D:

In a financial statement data extractor, runtime instrumentation observes actual proposed or completed tool activity, 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.

Question 557

Tests for a multi-repository troubleshooting agent will pass only if a patch that fails regression tests is diagnosed and corrected and if a destructive tool receives human authorization at execution time. Which proposed design directly satisfies both checks?

  1. rewrite large unrelated modules before re-running tests; and interrupt or deny the tool until an authorization callback approves it
  2. rewrite large unrelated modules before re-running tests; and collect a general approval at the start of the conversation
  3. inspect failure evidence and revise the smallest relevant change; and interrupt or deny the tool until an authorization callback approves it
  4. inspect failure evidence and revise the smallest relevant change; and collect a general approval at the start of the conversation

Explanation

Correct Answer

C:

In a multi-repository troubleshooting agent, test-driven feedback pinpoints a specific unmet behavior for the next iteration; alongside this, execution gates make the approval binding on the selected action and inputs, satisfying the distinct requirements that a patch that fails regression tests is diagnosed and corrected and that a destructive tool receives human authorization at execution time.

Incorrect Answers

A:

In a multi-repository troubleshooting agent, execution gates make the approval binding on the selected action and inputs, but broad edits create new regression surfaces while the cause remains unclear; therefore the separate requirement that a patch that fails regression tests is diagnosed and corrected would remain unsatisfied.

B:

In a multi-repository troubleshooting agent, broad edits create new regression surfaces while the cause remains unclear and blanket consent does not verify the later high-risk operation; the design would fail both conditions: a patch that fails regression tests is diagnosed and corrected and a destructive tool receives human authorization at execution time.

D:

In a multi-repository troubleshooting agent, test-driven feedback pinpoints a specific unmet behavior for the next iteration, 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 558

While preparing a static analysis automation for deployment, engineers must ensure that a destructive tool receives human authorization at execution time; they must also ensure that resuming a session does not bypass present-day authorization. Which combination best satisfies those requirements?

  1. interrupt or deny the tool until an authorization callback approves it; and recheck permissions and current resource state before new side effects
  2. collect a general approval at the start of the conversation; and treat old approval messages as unlimited authorization for later writes
  3. collect a general approval at the start of the conversation; and recheck permissions and current resource state before new side effects
  4. interrupt or deny the tool until an authorization callback approves it; and treat old approval messages as unlimited authorization for later writes

Explanation

Correct Answer

A:

In a static analysis automation, execution gates make the approval binding on the selected action and inputs; alongside this, conversation continuity is not a replacement for live access control or fresh tool state, satisfying the distinct requirements that a destructive tool receives human authorization at execution time and that resuming a session does not bypass present-day authorization.

Incorrect Answers

B:

In a static analysis automation, blanket consent does not verify the later high-risk operation and permissions and operational context can change between sessions; the design would fail both conditions: a destructive tool receives human authorization at execution time and resuming a session does not bypass present-day authorization.

C:

In a static analysis automation, conversation continuity is not a replacement for live access control or fresh tool state, 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.

D:

In a static analysis automation, execution gates make the approval binding on the selected action and inputs, but permissions and operational context can change between sessions; therefore the separate requirement that resuming a session does not bypass present-day authorization would remain unsatisfied.

Question 559

In a developer productivity assistant, the acceptance criteria require that independent research branches can finish without waiting on one another and that several requested client tools are returned in one response. Which paired implementation meets both?

  1. chain every research agent in a fixed serial pipeline; and execute every tool_use block and return a result with each corresponding ID
  2. dispatch the independent branches concurrently through the coordinator; and execute every tool_use block and return a result with each corresponding ID
  3. dispatch the independent branches concurrently through the coordinator; and execute only the first returned tool call
  4. chain every research agent in a fixed serial pipeline; and execute only the first returned tool call

Explanation

Correct Answer

B:

In a developer productivity assistant, parallel delegation reduces latency when individual investigations do not consume each other's outputs; alongside this, a multi-tool assistant turn can request independent actions and must receive all matching outcomes, satisfying the distinct requirements that independent research branches can finish without waiting on one another and that several requested client tools are returned in one response.

Incorrect Answers

A:

In a developer productivity assistant, a multi-tool assistant turn can request independent actions and must receive all matching outcomes, but unnecessary sequential handoffs increase latency for independent subtasks; therefore the separate requirement that independent research branches can finish without waiting on one another would remain unsatisfied.

C:

In a developer productivity assistant, parallel delegation reduces latency when individual investigations do not consume each other's outputs, but the unanswered tool uses remain unresolved and later reasoning loses evidence; therefore the separate requirement that several requested client tools are returned in one response would remain unsatisfied.

D:

In a developer productivity assistant, unnecessary sequential handoffs increase latency for independent subtasks and the unanswered tool uses remain unresolved and later reasoning loses evidence; the design would fail both conditions: independent research branches can finish without waiting on one another and several requested client tools are returned in one response.

Question 560

Before expanding a multi-source competitor research system, the owner must demonstrate that a high-risk result passes deterministic checks in addition to model critique, while also demonstrating that a delegated reviewer has no production write capability. Which design provides both controls?

  1. replace all validation checks with a second model's opinion; and provide only the read-oriented tools required by that reviewer
  2. combine independent review with schema and business-rule validation; and give the reviewer every administrative tool and ask for restraint
  3. combine independent review with schema and business-rule validation; and provide only the read-oriented tools required by that reviewer
  4. replace all validation checks with a second model's opinion; and give the reviewer every administrative tool and ask for restraint

Explanation

Correct Answer

C:

In a multi-source competitor research system, a critique may improve quality but cannot enforce machine-testable invariants by itself; alongside this, agent-scoped tool exposure prevents a write path in a read-only role, satisfying the distinct requirements that a high-risk result passes deterministic checks in addition to model critique and that a delegated reviewer has no production write capability.

Incorrect Answers

A:

In a multi-source competitor research system, agent-scoped tool exposure prevents a write path in a read-only role, but probabilistic agreement does not guarantee compliance; therefore the separate requirement that a high-risk result passes deterministic checks in addition to model critique would remain unsatisfied.

B:

In a multi-source competitor research system, a critique may improve quality but cannot enforce machine-testable invariants by itself, but prompt restraint is weaker than withholding execution capability; therefore the separate requirement that a delegated reviewer has no production write capability would remain unsatisfied.

D:

In a multi-source competitor research system, probabilistic agreement does not guarantee compliance and prompt restraint is weaker than withholding execution capability; the design would fail both conditions: a high-risk result passes deterministic checks in addition to model critique and a delegated reviewer has no production write capability.

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