Anthropic CCAR-P - Claude Certified Architect - Professional

The assistant is making recommendations without considering information that specialists regard as material. That is a direct input-scope and requirements gap, not merely a presentation issue. The 89 percent acceptance metric is also incomplete because initial acceptance does not reveal whether recommendations remain correct after users apply the omitted criteria. Production readiness requires outcome-focused evaluation, including final recommendation agreement, override reasons, financial impact, supplier risk, and performance across representative procurement scenarios. Options B and E raise possible concerns, but the scenario supplies no evidence that response rate or pilot duration is the demonstrated deficiency. Option A incorrectly treats a single proxy metric as sufficient. Anthropic recommends multidimensional, task-specific success criteria aligned with the actual user outcome. Define success criteria and evaluations

Option A augments the existing review process without transferring final accountability to the model. Claude Code can inspect the change in repository context, identify logic defects, security weaknesses, regressions, and maintainability concerns, and publish structured findings for human evaluation. Existing branch protection then preserves required approvals and separation of duties. Automatic merging or disabling protection rules would convert an advisory capability into an uncontrolled authorization mechanism. Silent deletion would eliminate transparency and prevent authors from challenging or correcting the assessment. Anthropic’s Code Review documentation states that findings are posted as comments and do not approve or block the pull request, leaving the established review workflow intact. Claude Code Review

Persist the validated output for downstream consumption and audit — Output Handling

Order prompt sections so cacheable content appears before per-request content — Prompt Construction

Summarize prior conversation history when full history is no longer needed — Input Preparation

Validate the model’s structured output against the expected schema — Output Handling

Move stable repeated content into a cacheable prefix at the beginning of the prompt — Prompt Construction

Trim retrieved passages to spans relevant to the user’s question — Input Preparation

Input preparation determines what evidence and history should enter the request, so summarization and retrieval trimming belong there. Prompt construction determines ordering and cache boundaries; stable repeated instructions must precede dynamic per-request content to maximize cache reuse. Output handling begins after generation and includes schema validation, persistence, auditing, and downstream delivery. Mixing these responsibilities creates inefficient prompts and weak validation boundaries. Anthropic’s context guidance emphasizes curating the smallest useful context because unnecessary tokens can reduce recall and accuracy. Prompt caching similarly depends on a stable reusable prefix, while structured outputs or schema validation protect downstream systems from malformed responses. Context windows; prompt caching

Haiku is the appropriate starting point because the workload is high-volume, latency-sensitive, and based on a stable closed set of moderation labels. These characteristics favor a fast, cost-efficient model capable of consistent classification without incurring the additional inference time and expense associated with deeper reasoning.

Anthropic’s model-selection guidance requires architects to balance capability, speed, and cost rather than automatically selecting the most capable model. Its content-moderation guidance specifically identifies Haiku as a cost-effective option for processing moderation workloads at substantial scale. Choosing the Right Model, Content Moderation

Opus is disproportionate to a routine closed-set classification problem. Sonnet may become justified if evaluation demonstrates that Haiku fails materially on complex policy distinctions, multilingual ambiguity, or adversarial edge cases, but it should not be selected merely because it is larger. Enabling extended thinking on every request would further increase latency and token consumption without evidence that the additional reasoning improves the defined success metrics. The correct architectural practice is to establish a representative moderation evaluation set, validate Haiku against accuracy and safety thresholds, and escalate only the cases that genuinely need deeper reasoning.

Study Guide references/topics: Model selection; capability–latency–cost trade-offs; classification workloads; evaluation-driven routing; moderation architecture.

===============

A new design cycle is warranted when the required change crosses an architectural boundary. Option D explicitly states that the existing architecture cannot satisfy the new requirements without redistributing component responsibilities or altering core contracts. Such changes may affect service ownership, orchestration, data flows, integration interfaces, security boundaries, failure handling, evaluation strategy, and operational accountability. They therefore require renewed discovery, impact analysis, design review, stakeholder approval, and regression planning.

Options A, B, and C are localized maintenance activities. Clarifying a runbook improves operational documentation but does not change the deployed system’s structure. Adjusting an alert threshold is a controlled operational tuning activity, provided the change remains within established monitoring policy. Editing customer-facing copy is similarly confined to the presentation or content layer. None of these changes inherently invalidates component contracts or architectural assumptions.

An architect should distinguish between iteration within an approved design and evidence that the design itself no longer supports the required outcome. Re-entering design for every minor adjustment creates unnecessary governance overhead; continuing local iteration after architectural assumptions have failed creates unmanaged technical and operational risk. Option D is the only signal that establishes a structural incompatibility.

Study Guide references/topics: Lifecycle feedback loops; design re-entry criteria; architectural significance; component responsibilities; interface contracts; controlled operational iteration.

===============

Control A directly constrains the outbound action surface. An allow-list limits which tools and destinations may receive data, while per-call review introduces an approval boundary before information leaves the controlled environment. These measures therefore address data exfiltration through outbound tool calls, making Option C correct. Anthropic’s security guidance recommends least privilege, narrowly scoped permissions, sandboxing, and limiting access to sensitive information and actions so that a compromised agent can cause minimal damage. Mitigate Jailbreaks and Prompt Injections

Control B addresses model-version drift. Running a stable adversarial evaluation set after each version change creates a consistent comparison baseline and detects regressions that might otherwise remain invisible during routine testing. This correctly maps Control B to silent quality drift, making Option E correct. Anthropic’s migration guidance repeatedly calls for workload reevaluation because instruction following, style, reasoning, and agent behavior can differ between versions. Model Migration Guide

Control A may limit the consequences of successful prompt injection, but it does not prevent or directly detect adversarial instructions in retrieved content. Control B measures behavioral performance but does not impose outbound authorization controls. Each control must therefore be mapped to the risk it most directly mitigates.

Study Guide references/topics: Tool governance; least privilege; exfiltration controls; adversarial evaluations; model-change regression testing; defense in depth.

===============

Option B uses progressive disclosure to meet the decision needs of both audiences. The executive sponsor first receives the selected direction, intended business outcomes, major constraints, cost or performance implications, and accepted trade-offs. This establishes why the decision matters and whether it remains aligned with strategic objectives.

The presentation can then move into the technical rationale required by engineering leads: component responsibilities, integration contracts, model and retrieval choices, security controls, rejected alternatives, implementation dependencies, evaluation gates, operational consequences, and migration requirements. This structure gives engineers enough evidence to implement the decision without forcing the executive audience to interpret unnecessary low-level detail.

Option A excludes the sponsor’s concerns and fails to connect technical design to business value. Option C removes the context required for governance and implementation, making the decision appear arbitrary. Option D assumes that all participants require identical information density, which usually produces a presentation that is too technical for executives and insufficiently precise for engineers.

The strongest communication pattern is layered but internally consistent: one decision, one rationale, and multiple levels of detail tailored to the responsibilities of the audience.

Study Guide references/topics: Mixed-audience communication; progressive disclosure; business outcomes; architectural trade-offs; technical rationale; implementation implications.

===============

Option D applies three complementary controls. First, a read-only database credential creates an authorization boundary outside the model; prompt instructions alone cannot convert that credential into write access. Second, restricting the subagent’s tool list reduces capability exposure by preventing the agent from selecting unrelated or unnecessarily privileged operations. Third, explicit human confirmation creates a deliberate approval gate before any exceptional data-changing action is executed.

This is defense in depth. If Claude misinterprets a request or processes malicious instructions from untrusted content, the restricted credential and tool configuration limit the available action surface. Human review then protects operations with potentially irreversible production consequences. Logging and audit trails should remain enabled to record the actor, request, tool call, approval, affected records, and outcome.

Anthropic documents that Claude Code begins with read-only permissions in Manual mode and requests approval for actions that modify the environment. Its permission system also supports granular allow, ask, and deny controls for MCP tools and subagents. Claude Code Security, Configure Permissions

Options A and B eliminate least privilege and approval boundaries. Option C removes essential detection and forensic evidence.

Study Guide references/topics: Production database safety; least privilege; read-only defaults; MCP permissions; subagent scoping; human confirmation; auditability.

===============