Question 100
Production readiness for a returns-and-billing assistant depends on ensuring that a corrected attempt preserves a record of prior validation findings and that source identity survives multi-agent research handoffs. Select the most appropriate combined approach.
- retain the original source, response version, and retry feedback; and carry citation identifiers and retrieval context through synthesis
- replace all earlier outputs without recording the validation history; and strip all citation fields before sending the writer its inputs
- retain the original source, response version, and retry feedback; and strip all citation fields before sending the writer its inputs
- replace all earlier outputs without recording the validation history; and carry citation identifiers and retrieval context through synthesis
Explanation
Correct Answer
A:In a returns-and-billing assistant, traceable attempts allow auditors to distinguish correction from silent overwrite; alongside this, provenance metadata must remain attached when intermediate summaries are merged, satisfying the distinct requirements that a corrected attempt preserves a record of prior validation findings and that source identity survives multi-agent research handoffs.
Incorrect Answers
B:In a returns-and-billing assistant, the reason for a changed value becomes untraceable and the writer cannot reliably reconstruct which source supports each conclusion; the design would fail both conditions: a corrected attempt preserves a record of prior validation findings and source identity survives multi-agent research handoffs.
C:In a returns-and-billing assistant, traceable attempts allow auditors to distinguish correction from silent overwrite, but the writer cannot reliably reconstruct which source supports each conclusion; therefore the separate requirement that source identity survives multi-agent research handoffs would remain unsatisfied.
D:In a returns-and-billing assistant, provenance metadata must remain attached when intermediate summaries are merged, but the reason for a changed value becomes untraceable; therefore the separate requirement that a corrected attempt preserves a record of prior validation findings would remain unsatisfied.
Question 101
Production readiness for an unfamiliar monorepo investigation depends on ensuring that a paused server-side tool loop can continue without losing work and that the repository assistant discovers filenames matching a pattern. Select the most appropriate combined approach.
- preserve the paused assistant response and resume the server-tool turn; and search entire file contents with Grep to find paths by naming pattern
- restart from a blank conversation whenever pause_turn occurs; and search entire file contents with Grep to find paths by naming pattern
- restart from a blank conversation whenever pause_turn occurs; and use Glob to locate candidate paths
- preserve the paused assistant response and resume the server-tool turn; and use Glob to locate candidate paths
Explanation
Correct Answer
D:In an unfamiliar monorepo investigation, pause_turn indicates an internal tool loop pause rather than a user-facing conclusion; alongside this, path globbing finds files by names and directory structure before content inspection, satisfying the distinct requirements that a paused server-side tool loop can continue without losing work and that the repository assistant discovers filenames matching a pattern.
Incorrect Answers
A:In an unfamiliar monorepo investigation, pause_turn indicates an internal tool loop pause rather than a user-facing conclusion, but content search answers a different question and may miss empty files; therefore the separate requirement that the repository assistant discovers filenames matching a pattern would remain unsatisfied.
B:In an unfamiliar monorepo investigation, dropping tool context can repeat work and omit intermediate evidence and content search answers a different question and may miss empty files; the design would fail both conditions: a paused server-side tool loop can continue without losing work and the repository assistant discovers filenames matching a pattern.
C:In an unfamiliar monorepo investigation, path globbing finds files by names and directory structure before content inspection, but dropping tool context can repeat work and omit intermediate evidence; therefore the separate requirement that a paused server-side tool loop can continue without losing work would remain unsatisfied.
Question 102
An architecture review of a shipping-exception helpdesk finds two missing guarantees: a corrected attempt preserves a record of prior validation findings; a transient upstream outage is retried safely. What two-part change closes both gaps?
- retain the original source, response version, and retry feedback; and apply a bounded backoff only to classified retryable errors
- replace all earlier outputs without recording the validation history; and apply a bounded backoff only to classified retryable errors
- replace all earlier outputs without recording the validation history; and retry access-denied and invalid-input errors indefinitely
- retain the original source, response version, and retry feedback; and retry access-denied and invalid-input errors indefinitely
Explanation
Correct Answer
A:In a shipping-exception helpdesk, traceable attempts allow auditors to distinguish correction from silent overwrite; alongside this, transient failures can heal without re-running permanently invalid requests, satisfying the distinct requirements that a corrected attempt preserves a record of prior validation findings and that a transient upstream outage is retried safely.
Incorrect Answers
B:In a shipping-exception helpdesk, transient failures can heal without re-running permanently invalid requests, but the reason for a changed value becomes untraceable; therefore the separate requirement that a corrected attempt preserves a record of prior validation findings would remain unsatisfied.
C:In a shipping-exception helpdesk, the reason for a changed value becomes untraceable and permanent failures require a fix or escalation rather than repetition; the design would fail both conditions: a corrected attempt preserves a record of prior validation findings and a transient upstream outage is retried safely.
D:In a shipping-exception helpdesk, traceable attempts allow auditors to distinguish correction from silent overwrite, but permanent failures require a fix or escalation rather than repetition; therefore the separate requirement that a transient upstream outage is retried safely would remain unsatisfied.
Question 103
Before expanding a customer refund workflow, the owner must demonstrate that repeated delivery of one action request cannot create duplicate writes, while also demonstrating that near-overlapping retrieval tools are selected by actual boundary. Which design provides both controls?
- use an idempotency key and confirm the prior commit before retrying; and rely exclusively on short tool names for routing
- use an idempotency key and confirm the prior commit before retrying; and document examples and conditions where each tool must not be used
- blindly resubmit the same write after every timeout; and document examples and conditions where each tool must not be used
- blindly resubmit the same write after every timeout; and rely exclusively on short tool names for routing
Explanation
Correct Answer
B:In a customer refund workflow, side-effecting operations need deduplication against timeouts and repeated messages; alongside this, contrastive guidance helps tool choice when resource names look similar, satisfying the distinct requirements that repeated delivery of one action request cannot create duplicate writes and that near-overlapping retrieval tools are selected by actual boundary.
Incorrect Answers
A:In a customer refund workflow, side-effecting operations need deduplication against timeouts and repeated messages, but names alone may not capture the distinction in scope or authorization; therefore the separate requirement that near-overlapping retrieval tools are selected by actual boundary would remain unsatisfied.
C:In a customer refund workflow, contrastive guidance helps tool choice when resource names look similar, but the first attempt might have committed despite a missing acknowledgment; therefore the separate requirement that repeated delivery of one action request cannot create duplicate writes would remain unsatisfied.
D:In a customer refund workflow, the first attempt might have committed despite a missing acknowledgment and names alone may not capture the distinction in scope or authorization; the design would fail both conditions: repeated delivery of one action request cannot create duplicate writes and near-overlapping retrieval tools are selected by actual boundary.
Question 104
In a technical due-diligence report generator, the acceptance criteria require that tool responses are associated with the exact invocations that produced them and that a delegated reviewer has no production write capability. Which paired implementation meets both?
- return tool results with newly invented identifiers; and give the reviewer every administrative tool and ask for restraint
- return tool_result blocks using the original tool_use_id values; and provide only the read-oriented tools required by that reviewer
- return tool results with newly invented identifiers; and provide only the read-oriented tools required by that reviewer
- return tool_result blocks using the original tool_use_id values; and give the reviewer every administrative tool and ask for restraint
Explanation
Correct Answer
B:In a technical due-diligence report generator, correlation identifiers bind asynchronous or parallel results to the right requests; alongside this, agent-scoped tool exposure prevents a write path in a read-only role, satisfying the distinct requirements that tool responses are associated with the exact invocations that produced them and that a delegated reviewer has no production write capability.
Incorrect Answers
A:In a technical due-diligence report generator, the model cannot reliably associate a result with an outstanding call and prompt restraint is weaker than withholding execution capability; the design would fail both conditions: tool responses are associated with the exact invocations that produced them and a delegated reviewer has no production write capability.
C:In a technical due-diligence report generator, agent-scoped tool exposure prevents a write path in a read-only role, but the model cannot reliably associate a result with an outstanding call; therefore the separate requirement that tool responses are associated with the exact invocations that produced them would remain unsatisfied.
D:In a technical due-diligence report generator, correlation identifiers bind asynchronous or parallel results to the right requests, 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.
Question 105
An audit of a repository release validation process asks how to make sure that a CI job invokes Claude without a terminal conversation. Which implementation should be selected?
- run Claude Code in non-interactive print mode with -p
- use the CLI structured-output schema option for the response
- run tests and use actual exit codes and artifacts for the pass decision
- launch an interactive Claude terminal in an unattended runner
Explanation
Correct Answer
A:For a repository release validation process, headless invocation permits scripted automation and exits after the requested task; this directly establishes that a CI job invokes Claude without a terminal conversation instead of relying on an unrelated workflow control.
Incorrect Answers
B:Although schema-constrained output is more robust for downstream integrations than prompt-only formatting, that only supports the distinct goal that an automation expects a defined JSON result schema, not the requirement that a CI job invokes Claude without a terminal conversation in a repository release validation process.
C:Although external deterministic checks prevent an assistant summary from declaring unexecuted tests successful, that only supports the distinct goal that a pipeline gate reflects objective test outcomes, not the requirement that a CI job invokes Claude without a terminal conversation in a repository release validation process.
D:In a repository release validation process, an interactive session can block automation waiting for a person; as a result, the stated requirement that a CI job invokes Claude without a terminal conversation would still not be met.
Question 106
Before expanding a merge-request policy gate, the owner must demonstrate that a small approved typo fix is completed without excessive planning, while also demonstrating that the CI parser receives predictable machine-readable Claude output. Which design provides both controls?
- launch a multi-stage planning effort for a single known line; and scrape a human-oriented terminal transcript with regular expressions
- use direct execution for the limited reviewed edit; and request the CLI JSON output format and parse its documented structure
- use direct execution for the limited reviewed edit; and scrape a human-oriented terminal transcript with regular expressions
- launch a multi-stage planning effort for a single known line; and request the CLI JSON output format and parse its documented structure
Explanation
Correct Answer
B:In a merge-request policy gate, a bounded low-risk change does not require a long exploratory workflow; alongside this, machine-readable envelopes avoid brittle extraction from colored console prose, satisfying the distinct requirements that a small approved typo fix is completed without excessive planning and that the CI parser receives predictable machine-readable Claude output.
Incorrect Answers
A:In a merge-request policy gate, extra ceremony adds latency without materially lowering risk and presentation formatting can change and break the parser; the design would fail both conditions: a small approved typo fix is completed without excessive planning and the CI parser receives predictable machine-readable Claude output.
C:In a merge-request policy gate, a bounded low-risk change does not require a long exploratory workflow, but presentation formatting can change and break the parser; therefore the separate requirement that the CI parser receives predictable machine-readable Claude output would remain unsatisfied.
D:In a merge-request policy gate, machine-readable envelopes avoid brittle extraction from colored console prose, but extra ceremony adds latency without materially lowering risk; therefore the separate requirement that a small approved typo fix is completed without excessive planning would remain unsatisfied.
Question 107
For a customer account helpdesk agent, acceptance depends on demonstrating that a paused server-side tool loop can continue without losing work. Which approach is the strongest fit?
- restart from a blank conversation whenever pause_turn occurs
- handle max_tokens as truncation rather than end_turn
- execute every tool_use block and return a result with each corresponding ID
- preserve the paused assistant response and resume the server-tool turn
- return tool_result blocks using the original tool_use_id values
Explanation
Correct Answer
D:For a customer account helpdesk agent, pause_turn indicates an internal tool loop pause rather than a user-facing conclusion; this directly establishes that a paused server-side tool loop can continue without losing work instead of relying on an unrelated workflow control.
Incorrect Answers
A:In a customer account helpdesk agent, dropping tool context can repeat work and omit intermediate evidence; as a result, the stated requirement that a paused server-side tool loop can continue without losing work would still not be met.
B:Although reaching an output limit is a stop condition but does not establish task completion, that only supports the distinct goal that a truncated response is not mistaken for successful completion, not the requirement that a paused server-side tool loop can continue without losing work in a customer account helpdesk agent.
C:Although a multi-tool assistant turn can request independent actions and must receive all matching outcomes, that only supports the distinct goal that several requested client tools are returned in one response, not the requirement that a paused server-side tool loop can continue without losing work in a customer account helpdesk agent.
E:Although correlation identifiers bind asynchronous or parallel results to the right requests, that only supports the distinct goal that tool responses are associated with the exact invocations that produced them, not the requirement that a paused server-side tool loop can continue without losing work in a customer account helpdesk agent.
Question 108
The release checklist for an analyst research coordinator explicitly checks that a partially failed batch is repaired efficiently; it also checks that an unresolved claim is qualified without inventing certainty. What coordinated solution meets the checklist?
- identify failed records and resubmit only safe retry candidates; and state a firm conclusion because the report needs a complete ending
- rerun the entire completed batch whenever one record fails; and express the uncertainty and identify what verification is missing
- rerun the entire completed batch whenever one record fails; and state a firm conclusion because the report needs a complete ending
- identify failed records and resubmit only safe retry candidates; and express the uncertainty and identify what verification is missing
Explanation
Correct Answer
D:In an analyst research coordinator, selective retry avoids paying again for already successful items; alongside this, calibrated language communicates a bounded evidence gap honestly, satisfying the distinct requirements that a partially failed batch is repaired efficiently and that an unresolved claim is qualified without inventing certainty.
Incorrect Answers
A:In an analyst research coordinator, selective retry avoids paying again for already successful items, but a polished narrative cannot supply missing corroboration; therefore the separate requirement that an unresolved claim is qualified without inventing certainty would remain unsatisfied.
B:In an analyst research coordinator, calibrated language communicates a bounded evidence gap honestly, but successful work would be duplicated unnecessarily; therefore the separate requirement that a partially failed batch is repaired efficiently would remain unsatisfied.
C:In an analyst research coordinator, successful work would be duplicated unnecessarily and a polished narrative cannot supply missing corroboration; the design would fail both conditions: a partially failed batch is repaired efficiently and an unresolved claim is qualified without inventing certainty.
Question 109
An incident postmortem for a service refactoring initiative requires controls so that a developer returns to an interrupted session with prior reasoning and that the assistant can use external service capabilities via MCP. Which paired remedy covers both failure modes?
- start an unrelated new session and assume it knows past actions; and configure a compatible MCP server and expose its declared tools to the client
- resume the stored session by its identifier; and configure a compatible MCP server and expose its declared tools to the client
- resume the stored session by its identifier; and describe the external API in a prompt without connecting any tool
- start an unrelated new session and assume it knows past actions; and describe the external API in a prompt without connecting any tool
Explanation
Correct Answer
B:In a service refactoring initiative, session resumption retains the recorded conversation and avoids rebuilding the entire task from scratch; alongside this, MCP bridges a client to provider-defined capabilities rather than inventing a Claude built-in, satisfying the distinct requirements that a developer returns to an interrupted session with prior reasoning and that the assistant can use external service capabilities via MCP.
Incorrect Answers
A:In a service refactoring initiative, MCP bridges a client to provider-defined capabilities rather than inventing a Claude built-in, but a fresh context lacks the prior interaction unless the facts are deliberately supplied; therefore the separate requirement that a developer returns to an interrupted session with prior reasoning would remain unsatisfied.
C:In a service refactoring initiative, session resumption retains the recorded conversation and avoids rebuilding the entire task from scratch, but instructions cannot create an executable integration endpoint; therefore the separate requirement that the assistant can use external service capabilities via MCP would remain unsatisfied.
D:In a service refactoring initiative, a fresh context lacks the prior interaction unless the facts are deliberately supplied and instructions cannot create an executable integration endpoint; the design would fail both conditions: a developer returns to an interrupted session with prior reasoning and the assistant can use external service capabilities via MCP.
Question 110
Tests for a contract clause extraction workflow will pass only if only one approved write tool is callable in a sensitive stage and if near-overlapping retrieval tools are selected by actual boundary. Which proposed design directly satisfies both checks?
- let the model select from a broad global list of tools; and document examples and conditions where each tool must not be used
- let the model select from a broad global list of tools; and rely exclusively on short tool names for routing
- restrict the stage's allowed tool set and verify permissions in the executor; and rely exclusively on short tool names for routing
- restrict the stage's allowed tool set and verify permissions in the executor; and document examples and conditions where each tool must not be used
Explanation
Correct Answer
D:In a contract clause extraction workflow, capability scoping narrows possible side effects before model selection; alongside this, contrastive guidance helps tool choice when resource names look similar, satisfying the distinct requirements that only one approved write tool is callable in a sensitive stage and that near-overlapping retrieval tools are selected by actual boundary.
Incorrect Answers
A:In a contract clause extraction workflow, contrastive guidance helps tool choice when resource names look similar, but excess capabilities create avoidable routing and authorization risk; therefore the separate requirement that only one approved write tool is callable in a sensitive stage would remain unsatisfied.
B:In a contract clause extraction workflow, excess capabilities create avoidable routing and authorization risk and names alone may not capture the distinction in scope or authorization; the design would fail both conditions: only one approved write tool is callable in a sensitive stage and near-overlapping retrieval tools are selected by actual boundary.
C:In a contract clause extraction workflow, capability scoping narrows possible side effects before model selection, but names alone may not capture the distinction in scope or authorization; therefore the separate requirement that near-overlapping retrieval tools are selected by actual boundary would remain unsatisfied.