Question 130
A code-audit skill requires helper scripts and reference examples that would distract every normal session. How should it be packaged?
- Split dependency installation behind explicit approval and documented prerequisites
- Specify required headings, evidence fields and a completion checklist
- Store supporting resources within the skill and load them when needed
- Explain the skill as plan analysis and deny deployment or mutation tools
Explanation
Correct Answer
C:Skills can bundle reference material and executable helpers for a focused task. On-demand loading keeps general conversation context lighter.
Incorrect Answers
A:Skills are reusable instructions, not a blanket permission grant for arbitrary software installation; its benefits for requiring approval for skill-driven dependency changes do not resolve the need for packaging on-demand skill resources.
B:A bounded output contract makes a reusable workflow easier to verify; supporting standardizing a command's output shape cannot replace the required packaging on-demand skill resources.
D:Skill language and access control should agree; that serves separating review skills from deployment skills, whereas the case requires packaging on-demand skill resources.
Question 131
An agent must find all files under migrations matching a suffix before evaluating upgrade order. Which tool sequence makes sense?
- Inspect the indicated path and relevant region directly
- Choose Read rather than invoking Bash to print the file
- Glob the matching migration paths and then inspect them in order
- Enforce a scoped Edit with an exact matching target
Explanation
Correct Answer
C:Filename enumeration establishes the migration set without executing anything. Subsequent reads can identify ordering information without risky command invocation.
Incorrect Answers
A:For avoiding unnecessary repository-wide discovery, a precise location is already supplied, so broad search adds noise and latency; however, this requirement concerns gathering migration files without executing code.
B:Reading file content through the dedicated tool avoids an unrestricted shell surface; that may serve choosing direct file inspection over shell execution, but it is no substitute for gathering migration files without executing code.
D:A targeted replacement can preserve neighboring configuration content; the stated problem requires gathering migration files without executing code rather than making a minimal configuration edit.
Question 132
A coordinator receives contradictory findings from two investigation subagents. What design preserves accountability?
- Provide each worker only the minimum task-specific context and authorized tools
- Include a targeted citation repair pass before further broad delegation
- Limit delegation depth and concurrent subagent count at the orchestration layer
- Resolve disagreements using traceable evidence and an explicit arbitration step
Explanation
Correct Answer
D:A coordinator should compare source support and record the unresolved variance rather than arbitrarily select whichever response arrived last. The arbitration point makes decisions auditable.
Incorrect Answers
A:Specialization should come with least-necessary information and access; the option would serve compartmentalizing delegated context and privileges, not reconciling conflicting worker conclusions as specified in the stem.
B:Missing source attribution is a bounded post-processing defect; this addresses targeted repair instead of repeated broad delegation, while the question is about reconciling conflicting worker conclusions.
C:Limits on nesting and worker breadth bound cost and latency; that relates to constraining recursive multi-agent delegation, not the necessary reconciling conflicting worker conclusions.
Question 133
A refund operation reports success, yet the account balance has not been rechecked. What closes the execution loop safely?
- Verify the side effect with an authoritative follow-up read before promising completion
- Provide the completed assistant message to the caller
- Link bounded retries with an operation idempotency key and outcome lookup
- Pair and execute both tool calls, returning their respective results
Explanation
Correct Answer
A:A tool's acknowledgement can be stale or incomplete relative to durable state. Checking the system of record before a definitive customer promise improves reliability around externally visible actions.
Incorrect Answers
B:The end_turn reason indicates the model finished its turn; the orchestration loop can surface the produced answer rather than manufacturing another tool step; the unsatisfied requirement is verifying externally visible side effects, not recognizing a completed model turn.
C:Retries of side-effecting actions can duplicate work when acknowledgment is lost; in this scenario, safe recovery after an ambiguous tool timeout is secondary to verifying externally visible side effects.
D:When considering handling parallel client-tool calls, a single model response can request multiple client tools; the host should process each authorized call and retain its unique tool-use identifier so results cannot be crossed; the requirement for verifying externally visible side effects remains unmet.
Question 134
An extraction process begins to retry every failure, including authorization denials from the storage API. What should change?
- Collapse repeated entries in extracted records using source identifiers and document context
- Require a better source or human review instead of repeated identical retries
- Forward uncertain critical addresses to a manual exception review
- Classify failure types and retry only plausibly transient conditions
Explanation
Correct Answer
D:Authentication and permission errors require configuration or credential repair. Treating them as transient wastes attempts and hides the underlying access problem.
Incorrect Answers
A:Overlapping page captures can lead to double counting; that addresses removing duplicate extraction artifacts, whereas this case calls for distinguishing retryable and nonretryable failures.
B:An absent source page cannot be recovered by sampling the same incomplete input; this concerns recognizing a nonrecoverable extraction input gap, leaving the requirement for distinguishing retryable and nonretryable failures outstanding.
C:For shipping or compliance, uncertain addresses have operational consequences; routing low-confidence addresses to human review and distinguishing retryable and nonretryable failures are distinct responsibilities under these constraints.
Question 135
A reviewer finds that a cited web page changed after a report was published. What archival practice supports traceability?
- Record retrieval time and a source version or approved snapshot reference
- Classify confirmed observations separately from estimates and unknowns
- Retain original-source identifiers and evidence type through every handoff
- Check that every cited claim has an inspected supporting source
Explanation
Correct Answer
A:Live pages can change or disappear. Time-stamped provenance helps investigators understand which evidence was available when the decision was made.
Incorrect Answers
B:Provenance distinguishes system-of-record data from inference; that serves separating verified status from estimates, whereas the case requires preserving time-sensitive research provenance.
C:Because a claim's evidentiary strength depends on where it originated, the measure supports retaining primary-source lineage across agents rather than the needed preserving time-sensitive research provenance.
D:A plausible citation string is not evidence of source support; its benefits for checking citations against actually inspected evidence do not resolve the need for preserving time-sensitive research provenance.
Question 136
A supervisor approves a recommendation, but the agent's workflow later changes the action amount. What approval rule is needed?
- Maintain individual rationales, evidence and the final adjudication basis
- Insist on evidence and calibration rather than trusting the self-reported confidence number
- Expose policy basis, supporting records, and the requested action before approval
- Bind approval to immutable action parameters and require reapproval after material changes
Explanation
Correct Answer
D:Approval of one transaction should not authorize a modified one. Parameter binding protects against drift between review and execution.
Incorrect Answers
A:Disagreement may expose ambiguous policies or missing facts; this may support retaining review disagreement and adjudication provenance but cannot alone meet the requirement for tying human approval to exact mutation details.
B:A model's expressed confidence may not correspond to measured probability; the stated problem requires tying human approval to exact mutation details rather than avoiding uncalibrated confidence as proof.
C:Meaningful approval requires adequate context to judge eligibility and consequences; a method suited to giving human reviewers sufficient decision evidence does not meet the need for tying human approval to exact mutation details.
Question 137
The model must return a conversational answer without calling any external tool for a greeting. How should tool use be configured?
- Provide separated tool sets or filtered capabilities according to worker identity
- Focus on limiting filesystem-capable tools independently from remote read-only search tools
- Implement a scoped tool allowlist for the worker invocation
- Allow the model to answer without an obligatory tool invocation
Explanation
Correct Answer
D:Optional tool selection prevents unnecessary calls when the request requires no external data. Forcing a call in these cases raises cost and side-effect risk without adding information.
Incorrect Answers
A:Different trust levels deserve different callable surfaces; that relates to partitioning sensitive and ordinary MCP capabilities, not the necessary avoiding forced calls on tool-free requests.
B:Although both operations may look like information retrieval, but local file access can reveal credentials, the choice addresses separating local file access from remote search and not the requested avoiding forced calls on tool-free requests.
C:A narrow allowance reduces both accidental mutation and prompt-injection opportunities; the deciding condition is avoiding forced calls on tool-free requests, whereas this action concerns applying per-worker tool allowlists.
Question 138
A contract extraction is missing a signature that is legally necessary for downstream filing. Which path should the system take?
- Resolve the intended action and confirm authorization before choosing either operation
- Transfer the case to a human approver with amount and policy evidence
- Flag the missing signature and request document correction or human review
- Require policy-owner resolution before applying either disputed instruction
Explanation
Correct Answer
C:Critical missing evidence cannot be filled from a language model's guess. The workflow should stop the regulated downstream action until the source is complete.
Incorrect Answers
A:Destructive and reversible operations are not interchangeable; in this scenario, clarifying an ambiguous destructive request is secondary to escalating missing legally required evidence.
B:When considering routing above-limit refunds to a human, threshold exceptions require explicit authority; the requirement for escalating missing legally required evidence remains unmet.
D:Conflicting authoritative sources create genuine uncertainty; escalating conflicting governing policies and escalating missing legally required evidence require different controls in this situation.
Question 139
A software review agent is asked to inspect generated patches and also apply corrections instantly. What separation is prudent?
- Have the reviewer propose evidence-backed findings and require authorized implementation afterward
- Sort findings by impact and evidence strength before producing recommendations
- Perform an independent verifier against the original requirements and observed tool evidence
- Contrast each candidate value to document evidence and escalate unresolved disagreement
Explanation
Correct Answer
A:Reviewer and implementer responsibilities differ. Keeping mutation outside the review pass prevents unapproved changes and makes findings independently auditable.
Incorrect Answers
B:Review output should prioritize well-supported important issues; the action supports prioritizing findings after independent review but falls short of separating review conclusions from edit authority in this case.
C:Self-assessment tends to reuse the model's assumptions; this can serve reducing self-confirmation bias in generated artifacts, but leaves the requirement for separating review conclusions from edit authority unmet.
D:Model disagreement does not supply a factual tie-breaker; that addresses resolving extraction disagreements with evidence, whereas this case calls for separating review conclusions from edit authority.
Question 140
A quality-review skill contains hard-coded company secrets in its reference file. What is the correct fix?
- Save supporting resources within the skill and load them when needed
- Set up isolated or forked context for the specialized skill
- Remove embedded secrets and access approved credentials through secure runtime configuration
- Establish a custom slash-invocable command or skill for the PR summary
Explanation
Correct Answer
C:Skill contents may be loaded into model context and shared in repositories. Credentials need separate guarded provisioning and should be rotated if exposed.
Incorrect Answers
A:Because skills can bundle reference material and executable helpers for a focused task, the measure supports packaging on-demand skill resources rather than the needed securing reusable skill contents.
B:Context isolation can keep reviewer instructions and intermediate analysis apart from the parent task; its benefits for isolating a specialized skill's context do not resolve the need for securing reusable skill contents.
D:Named command workflows provide repeatable behavior for a familiar task; this contributes to creating a reusable named PR-summary action without meeting the need for securing reusable skill contents.