Question 120
The defining problem for a field-service coordination pilot is that primary inference service can time out during a critical support window. Which proposed approach addresses it most precisely?
- Implement Context-capacity fit before model selection
- Implement Fallback model under failure
- Implement Empirical model bake-off
- Implement Reasoning-capable model for complex synthesis
Explanation
Correct Answer
B:The mechanism is specific: the model configuration can route to an approved fallback model with compatible output contracts. A credible implementation requires model contract compatibility tests and outage policy, while its acceptance test monitors safe fallback completion rate under simulated outages. The alternative approaches leave the risk that failover to an untested model can change schema and decision quality.
Incorrect Answers
A:When legal analysis requests require unusually long source packets, the model configuration can test effective context utilization and recall before choosing a model configuration; that does not provide the capability actually needed: the model configuration can route to an approved fallback model with compatible output contracts.
C:If stakeholders choose a model based on brand rather than target workload, the model configuration can evaluate candidate models on identical task data and operating constraints; whereas the field-service coordination case must guard against the outcome that failover to an untested model can change schema and decision quality.
D:If multi-document arbitration requires several interdependent inferences, a stronger model can justify tradeoffs across heterogeneous evidence; that does not provide the capability actually needed: the model configuration can route to an approved fallback model with compatible output contracts.
Question 121
A contract administration team is selecting the next improvement for its AI service; executives claim the assistant will save time without measuring current work. What design choice has the strongest direct justification?
- Adopt Outcome-first use-case framing
- Adopt Baseline workflow benchmarking
- Adopt Value-to-risk prioritization
- Adopt Feasibility pilot with representative cases
- Adopt Task-scope exclusion for unsuitable decisions
Explanation
Correct Answer
B:The proposed control fits because measuring existing effort establishes a credible incremental-value estimate. Its production prerequisites include a baseline observation period and matching task definitions; check end-to-end labor minutes per accepted outcome on representative tasks. Without this boundary, deployment can appear successful while adding hidden review effort.
Incorrect Answers
A:When stakeholders disagree whether faster drafting or fewer escalations is the goal, a measurable business outcome narrows the solution boundary before model selection; the current decision instead depends on a specific capability: measuring existing effort establishes a credible incremental-value estimate.
C:If candidate use cases differ in savings, uncertainty and personal-data exposure, a scored decision framework surfaces high-value lower-risk starting points; the missing operational behavior is specific: measuring existing effort establishes a credible incremental-value estimate.
D:When the proposed automation encounters exceptions that have never been tested, small-scope trials expose feasibility and exception handling before a broad commitment; the current decision instead depends on a specific capability: measuring existing effort establishes a credible incremental-value estimate.
E:When a client demands a guarantee of fully accurate autonomous legal advice, identifying out-of-scope judgments prevents unsupported automation promises; whereas the contract administration case must guard against the outcome that deployment can appear successful while adding hidden review effort.
Question 122
Following a regression in public-sector casework, the team establishes that unrestricted execution can alter secrets or infrastructure accidentally. Which proposed redesign most directly prevents recurrence?
- Deploy Repository-level project instructions
- Deploy Least-privilege tool permissions
- Deploy Hook-based policy checks
- Deploy Reproducible development environment
Explanation
Correct Answer
B:This intervention addresses the failure at its source because the engineering workflow can configure approved actions and deny risky operations by default. To keep the issue from recurring, establish team permission policy and escalation route and verify blocked unsafe command attempts in controlled tests; otherwise unrestricted execution can alter secrets or infrastructure accidentally can persist.
Incorrect Answers
A:When personal chat instructions do not establish reproducible team conventions, the engineering workflow can version shared Claude Code guidance alongside the project and review changes; but this remedy operates on another boundary, not the case where unrestricted execution can alter secrets or infrastructure accidentally.
C:When natural-language reminders cannot substitute for enforceable test execution, the engineering workflow can run deterministic checks at controlled workflow boundaries; yet it would not correct the observed weakness: unrestricted execution can alter secrets or infrastructure accidentally.
D:Where agent-assisted fixes work on one laptop but fail in CI, the engineering workflow can pin dependencies, tooling and environment assumptions across contributors; but this remedy operates on another boundary, not the case where unrestricted execution can alter secrets or infrastructure accidentally.
Question 123
For a contract administration rollout, the proposed design uses Unsafe-action escape rate, relying on the claim that the evaluation process can measure disallowed actions that reach a protected side-effecting endpoint. What enabling control must be in place to make that design enforceable?
- Confirm reference documents and a citation validation rubric
- Confirm complete cost attribution and ticket acceptance rules
- Confirm gold labels including unanswerable requests
- Confirm policy test harness and action audit events
Explanation
Correct Answer
D:The prerequisite creates the necessary operating boundary: policy test harness and action audit events. The system is intended to work because the evaluation process can measure disallowed actions that reach a protected side-effecting endpoint. A release without that foundation may still suffer general answer accuracy can hide rare severe tool abuse; use unauthorized side effects per adversarial test run to evaluate the result.
Incorrect Answers
A:The setup is useful to enable the alternative's behavior where the evaluation process can score whether factual claims are supported by cited source spans and the system summarizes regulated policies with citations; for this release, the team must first verify policy test harness and action audit events.
B:The setup is useful to enable the alternative's behavior where the evaluation process can include review, retries and inference in outcome-based unit economics and model usage drops while human rework unexpectedly grows; however, the named intervention needs policy test harness and action audit events before its protection can be enforced.
C:This prerequisite supports a method used when incomplete evidence should yield escalation rather than invented certainty, because the evaluation process can evaluate correctness jointly with justified abstentions across uncertainty levels; but the production dependency for Unsafe-action escape rate is policy test harness and action audit events.
Question 124
The architecture review for human-resources services has this operational target: the integration can expose read-only access independently from privileged changes. Which THREE choices provide the solution, its enabling condition, and a defensible release measure? Choose THREE.
- Implement Read-write capability separation
- Establish distinct service identities and audited write approvals
- Track prohibited-call rejection and schema-compliant invocation rates
- Establish inventory with owner, version and retirement process
- Track successful read tasks without elevated write privilege
- Implement Capped tool-loop execution
Explanation
Correct Answers
A:Implementing the intervention changes the operational mechanism so that the integration can expose read-only access independently from privileged changes. The failure to prevent is that mixing reading and mutation in one tool overgrants routine sessions; configuration and validation remain separate release obligations.
B:Before rollout, the team needs distinct service identities and audited write approvals. This is a deployment dependency for the intervention whose effect is that the integration can expose read-only access independently from privileged changes, not an alternative metric or unrelated design feature.
E:This is an acceptance signal rather than another design control: successful read tasks without elevated write privilege. Evaluating that signal on representative tasks tests whether the integration can expose read-only access independently from privileged changes and identifies cases where mixing reading and mutation in one tool overgrants routine sessions.
Incorrect Answers
C:This test would validate a different benefit: the integration can replace broad verbs with narrow parameterized operations and validation; but that indicator does not verify successful read tasks without elevated write privilege for the selected architecture.
D:The alternative setup enables a different mechanism: the integration can retire or consolidate redundant capabilities with a responsible owner; whereas this intervention requires distinct service identities and audited write approvals as its operating precondition.
F:This design choice is aimed at cases where a search agent endlessly retries ineffective tools on malformed identifiers; it is effective because the integration can enforce retry, token and call budgets with a deliberate stop path. The rollout described is instead governed by the case where an assistant must read audit events but rarely modify audit settings and requires a different primary intervention; however, the intended intervention requires that the integration can expose read-only access independently from privileged changes.
Question 125
An operations analyst for legal operations must evaluate Current-state service blueprint on real tasks. A test must detect whether redesigning a guessed process can shift work rather than remove it, even when outputs appear fluent. Which measure is appropriate?
- Track agreed priority scores and expected benefits for shortlisted uses
- Track proportion of nonfunctional requirements with acceptance tests
- Track discovered exceptions covered by the agreed requirements
- Track baseline handoff count and exception cycle time
Explanation
Correct Answer
D:The correct signal is baseline handoff count and exception cycle time; it must be computed for representative workloads and compared with a baseline. It checks whether the delivery team can chart the existing input-to-outcome process and exception volumes while surfacing the exposure that redesigning a guessed process can shift work rather than remove it.
Incorrect Answers
A:The proposed measure would suit a setting where teams propose many attractive pilots without common selection criteria, since the delivery team can jointly rank candidates by value, feasibility, harm and data readiness; by contrast, the public-sector casework release needs evidence from baseline handoff count and exception cycle time.
B:When a proposal lists features but omits response time, recovery requirements and the delivery team can capture distinct behavior, latency, security, availability and data constraints; however, that proxy would not demonstrate baseline handoff count and exception cycle time for the selected intervention.
C:When sponsors describe an ideal process that differs from daily operations, the delivery team can observe actual exception handling and frontline work before mapping automation; but the acceptance gate must instead inspect baseline handoff count and exception cycle time to detect the targeted failure.
Question 126
A public-sector casework pilot has enabled Rate and spend limits by tenant. Because a global cost cap allows one tenant to crowd out others, the release board wants evidence of effectiveness, not just a favorable demonstration. Which metric most directly tests the proposed control?
- Monitor attack containment time and residual failure severity
- Monitor cross-tenant fairness and quota bypass incident counts
- Monitor unapproved high-risk actions prevented during simulations
- Monitor sensitive data leakage rate in response audits
Explanation
Correct Answer
B:A defensible evaluation looks at cross-tenant fairness and quota bypass incident counts. The control's causal claim is that the control layer can enforce bounded usage budgets and anomaly controls per tenant. Monitoring that outcome, not merely elapsed tokens or success-looking text, helps catch a global cost cap allows one tenant to crowd out others.
Incorrect Answers
A:Where failures may bypass one input or model safety layer, a relevant check would establish whether the control layer can combine prevention, detection, containment and audited recovery; however, that proxy would not demonstrate cross-tenant fairness and quota bypass incident counts for the selected intervention.
C:Where proposed external transfers could produce irreversible customer harm, a relevant check would establish whether the control layer can stop before execution and require authorized human confirmation; by contrast, the public-sector casework release needs evidence from cross-tenant fairness and quota bypass incident counts.
D:When generated summaries may echo sensitive identifiers unnecessary for recipients, the control layer can enforce recipient-specific redaction and validated output fields; however, that proxy would not demonstrate cross-tenant fairness and quota bypass incident counts for the selected intervention.
Question 127
An insurance processing implementation team has approved Baseline workflow benchmarking. Its expected behavior is that measuring existing effort establishes a credible incremental-value estimate. What prerequisite belongs on the release-readiness checklist?
- Verify a documented decision authority and approval queue
- Verify a joint rubric agreed with security and operations
- Verify hypotheses with owners and stage-gate thresholds
- Verify a baseline observation period and matching task definitions
Explanation
Correct Answer
D:Readiness depends on a baseline observation period and matching task definitions, not just a design approval. That condition supports the mechanism in which measuring existing effort establishes a credible incremental-value estimate. Without it, deployment can appear successful while adding hidden review effort; monitoring end-to-end labor minutes per accepted outcome helps verify the deployed result.
Incorrect Answers
A:Where caseworkers must retain authority for consequential final decisions, Claude drafts ranked recommendations while a designated reviewer authorizes each action; however, the named intervention needs a baseline observation period and matching task definitions before its protection can be enforced.
B:The setup is useful to enable the alternative's behavior where a scored decision framework surfaces high-value lower-risk starting points and candidate use cases differ in savings, uncertainty and personal-data exposure; but the production dependency for Baseline workflow benchmarking is a baseline observation period and matching task definitions.
C:For the separate challenge where a team wants to fund a large platform before validating which use case helps, staged hypotheses tie each funding tranche to measured evidence; for this release, the team must first verify a baseline observation period and matching task definitions.
Question 128
The compliance assurance go-live board accepts the premise that the model configuration can route to an approved fallback model with compatible output contracts. Which preparation is indispensable before it can rely on Fallback model under failure operationally?
- Arrange representative examples and equalized test settings
- Arrange long-context benchmark cases and token accounting
- Arrange model contract compatibility tests and outage policy
- Prepare benchmark cases with evidence spanning multiple sources
Explanation
Correct Answer
C:The required preparation is model contract compatibility tests and outage policy. It makes the following architecture claim enforceable: the model configuration can route to an approved fallback model with compatible output contracts. If that setup is skipped, failover to an untested model can change schema and decision quality, so the release should examine safe fallback completion rate under simulated outages.
Incorrect Answers
A:For the separate challenge where stakeholders choose a model based on brand rather than target workload, the model configuration can evaluate candidate models on identical task data and operating constraints; but the production dependency for Fallback model under failure is model contract compatibility tests and outage policy.
B:For the separate challenge where legal analysis requests require unusually long source packets, the model configuration can test effective context utilization and recall before choosing a model configuration; for this release, the team must first verify model contract compatibility tests and outage policy.
D:This prerequisite supports a method used when multi-document arbitration requires several interdependent inferences, because a stronger model can justify tradeoffs across heterogeneous evidence; but the production dependency for Fallback model under failure is model contract compatibility tests and outage policy.
Question 129
An energy operations service owner is reviewing Unsafe-action escape rate after a canary rollout. The critical regression would occur if general answer accuracy can hide rare severe tool abuse. What observable acceptance signal belongs in the dashboard?
- Audit high-percentile latency and timeout frequency
- Audit total spend divided by validated resolved tickets
- Audit risk-coverage curve and improper certainty rate
- Audit unauthorized side effects per adversarial test run
Explanation
Correct Answer
D:The acceptance measure focuses on unauthorized side effects per adversarial test run at the task or workflow boundary. Model fluency would not establish whether the evaluation process can measure disallowed actions that reach a protected side-effecting endpoint, and it could conceal cases where general answer accuracy can hide rare severe tool abuse.
Incorrect Answers
A:The alternative indicator checks whether the evaluation process can track p95 and p99 end-to-end latency by request class when interactive sessions frequently exceed their response deadline; however, that proxy would not demonstrate unauthorized side effects per adversarial test run for the selected intervention.
B:When model usage drops while human rework unexpectedly grows, the evaluation process can include review, retries and inference in outcome-based unit economics; but the acceptance gate must instead inspect unauthorized side effects per adversarial test run to detect the targeted failure.
C:When incomplete evidence should yield escalation rather than invented certainty, the evaluation process can evaluate correctness jointly with justified abstentions across uncertainty levels; by contrast, the energy operations release needs evidence from unauthorized side effects per adversarial test run.
Question 130
The defining problem for a travel operations pilot is that a search agent endlessly retries ineffective tools on malformed identifiers. Which proposed approach addresses it most precisely?
- Introduce Capped tool-loop execution
- Introduce Single-purpose tool contract
- Introduce Read-write capability separation
- Introduce Progressive tool discovery
Explanation
Correct Answer
A:Under the stated constraints, the integration can enforce retry, token and call budgets with a deliberate stop path. That directly addresses the operational failure without relying on a generic prompt reminder. Production depends on loop policy and classified tool-error responses; the result to monitor is attempts per failed case and bounded-failure completion rate.
Incorrect Answers
B:For a workload where a generic execute-any-query tool lets a model submit unreviewed commands, the integration can replace broad verbs with narrow parameterized operations and validation; whereas the travel operations case must guard against the outcome that uncontrolled retries exhaust quotas and conceal a failed precondition.
C:For a workload where an assistant must read audit events but rarely modify audit settings, the integration can expose read-only access independently from privileged changes; that does not provide the capability actually needed: the integration can enforce retry, token and call budgets with a deliberate stop path.
D:When a complex connector ecosystem cannot fit full tool descriptions in each request, the integration can advertise compact discovery metadata and load detailed definitions on demand; the alternative cannot replace this key requirement: the integration can enforce retry, token and call budgets with a deliberate stop path.