Free CCAR-P Developer Productivity & Operational Enablement Practice Questions

The Developer Productivity & Operational Enablement domain makes up 7% of the Anthropic CCAR-P exam. Practice free questions covering developer tooling, Claude Code enablement, operational runbooks, monitoring, and team productivity practices — each with the correct answer and a detailed explanation. The real exam mixes multiple-choice and multiple-response items; these practice questions use a single-best-answer format.

CCAR-P Developer Productivity & Operational Enablement Sample Questions with Answers

Sample Question 1 — Developer Productivity & Operational Enablement

A platform team has 45 developers contributing to the same monorepo. Developers using Claude-assisted tooling repeatedly provide different build commands, test locations, and pull-request requirements. This has caused generated changes to bypass required integration tests, and new developers need several days to learn the conventions. Some developers also want to retain personal formatting and explanation preferences. Which configuration approach should the architect recommend?

  1. A. Let each developer maintain a complete local instruction file and distribute an example through the onboarding wiki.
  2. B. Store mandatory repository workflows in reviewed, version-controlled project instructions and keep optional personal preferences in user-specific configuration. (Correct answer)
  3. C. Embed all repository and personal instructions in a wrapper script maintained outside the repository by the platform team.
  4. D. Require developers to paste the current contribution guide into every session and review generated changes for missing steps.

Correct answer: B

Explanation: Correct answer (B): The shared, version-controlled instructions establish one durable source for mandatory build, test, and review conventions, while user configuration preserves harmless personalization. The governing constraint is reducing inconsistent changes without eliminating justified individual preferences. The architectural principle is to separate centrally governed project requirements from user-specific settings and manage shared AI configuration as an engineering asset. This matters in production because reviewed, traceable configuration reduces drift, accelerates onboarding, and makes changes to development behavior auditable. Why the other options are wrong: - Option A: Local files support customization, but duplicating mandatory conventions across developers preserves the configuration drift causing the current failures. - Option C: A wrapper can standardize execution, but keeping the instructions outside the repository weakens visibility, version alignment, and normal code-review governance. - Option D: Repeatedly supplying documentation may improve individual sessions, but it creates manual overhead and does not reliably establish a consistent shared baseline.

Sample Question 2 — Developer Productivity & Operational Enablement

A payments company is enabling Claude-assisted development in a repository that contains deployment scripts. The current environment gives the assistant unrestricted shell access and inherits each developer's production credentials. Developers need Claude to inspect source files, edit a working branch, run approved local tests, and query a sanitized test database. Security requires structural prevention of production changes and secret exposure; retaining a command log alone is not sufficient. Which design BEST meets these requirements?

  1. A. Keep unrestricted shell access, add instructions prohibiting production commands, and alert security when sensitive command patterns appear.
  2. B. Run sessions in an isolated workspace with scoped file access, allowlisted test commands, test-only credentials, and no production deployment capability. (Correct answer)
  3. C. Permit read-only production access, block file writes by default, and let developers temporarily enable unrestricted access when implementation begins.
  4. D. Retain developer credentials but require explicit confirmation before every shell command and archive complete session transcripts for audit.

Correct answer: B

Explanation: Correct answer (B): An isolated workspace with narrow file, command, and credential capabilities directly removes the paths to production modification and secret exposure while preserving the required development tasks. The decisive constraint is deterministic prevention rather than detection after an unsafe command. The architectural principle is least privilege enforced at the tool and identity boundaries, not through prompts or logs. This matters because model or user mistakes cannot exercise capabilities that the environment never grants. Why the other options are wrong: - Option A: Prompt restrictions and alerts are useful compensating controls, but unrestricted access leaves the underlying production-change capability available. - Option C: Separating read and write access can reduce risk, but temporary unrestricted access violates the requirement to prevent production access structurally. - Option D: Confirmations and transcripts improve oversight and attribution, but inherited production credentials still expose unnecessary high-impact capabilities and secrets.

Sample Question 3 — Developer Productivity & Operational Enablement

A manufacturer plans to use Claude to modernize a 220,000-line order-allocation service. Documentation is incomplete, test coverage is 28%, and several undocumented allocation rules affect contractual delivery commitments. The replacement must remain behaviorally compatible, and releases need a reversible path. Management wants useful increments within three months rather than a prolonged analysis phase. Which delivery workflow should the architect recommend?

  1. A. Characterize critical behavior with executable tests, map dependencies, modernize bounded slices, and require regression gates and developer approval for each increment. (Correct answer)
  2. B. Generate a full replacement from the existing source, compare API schemas, and deploy it once both implementations compile without warnings.
  3. C. Expand unit coverage for the entire legacy service before changing code, then perform the modernization as one reviewed release.
  4. D. Run the generated replacement beside production, compare aggregate outputs for one week, and automatically redirect all traffic when agreement is high.

Correct answer: A

Explanation: Correct answer (A): Characterization tests capture undocumented contractual behavior, while bounded increments provide early progress, isolate regressions, and support rollback. The interacting constraints are behavioral compatibility, poor existing coverage, a three-month demand for useful increments, and reversibility. The governing principle is to decompose risky modernization into inspectable stages with executable evidence and human accountability. In production, this limits blast radius and prevents compilation or aggregate agreement from being mistaken for semantic equivalence. Why the other options are wrong: - Option B: A generated replacement may accelerate implementation, but compilation and schema compatibility do not validate undocumented allocation behavior. - Option C: Comprehensive test expansion would reduce uncertainty, but covering the entire service before delivery conflicts with the need for useful increments within three months. - Option D: Parallel comparison can provide production evidence, but aggregate agreement may hide rare contractual cases and automatic cutover lacks the required controlled, reversible progression.

Sample Question 4 — Developer Productivity & Operational Enablement

A product organization expects Claude to create approximately 300 pull requests per week. Unit tests currently pass on 96% of generated changes, but a pilot found authorization defects and incompatible dependency updates in several passing pull requests. The organization wants faster delivery while retaining accountable ownership for security-sensitive code. Which pull-request policy BEST balances these requirements?

  1. A. Automatically merge every pull request that passes unit tests, then use production monitoring and automated rollback to address escaped defects.
  2. B. Require identical manual security and architecture review for every generated pull request, regardless of affected files or change type.
  3. C. Apply automated unit, integration, static-analysis, and dependency gates, then route changes to risk-based human review using code ownership. (Correct answer)
  4. D. Ask Claude to critique each generated change, merge when its critique reports no concerns, and sample ten percent for human review.

Correct answer: C

Explanation: Correct answer (C): Layered automated gates address distinct failure classes, while code ownership and risk-based review preserve human accountability where consequences are higher. The key evidence is that unit-test success has not detected authorization and dependency defects, yet reviewing every change identically would undermine the throughput objective. The architectural principle is proportionate oversight supported by independent validation. In production, this improves delivery speed without treating a single test layer or model self-review as sufficient assurance. Why the other options are wrong: - Option A: Automated merging and rollback can suit low-risk reversible changes, but authorization defects may expose data before monitoring detects them. - Option B: Universal manual review provides oversight, but equal treatment of all changes ignores risk and unnecessarily limits the requested throughput improvement. - Option D: Model critique can help reviewers, but it is not independent enforcement and sampled review does not preserve ownership for every security-sensitive change.

Sample Question 5 — Developer Productivity & Operational Enablement

An operations team is introducing a Claude-based assistant for severity-one incidents. It must inspect logs, traces, deployment records, source code, and current configuration. During incidents, evidence is incomplete and an incorrect command could extend an outage affecting 80,000 users. Existing production changes require an approved operator and a documented rollback plan. What should the assistant's initial capability boundary be?

  1. A. Provide read-only diagnostic tools and require approved operators to execute proposed remediation through the existing change process. (Correct answer)
  2. B. Provide unrestricted production shell access but require the assistant to explain each command before running it.
  3. C. Allow automatic rollback of any recent deployment while keeping all other production actions read-only.
  4. D. Allow configuration writes during severity-one incidents and rely on command logging for post-incident accountability.

Correct answer: A

Explanation: Correct answer (A): Read-only diagnostic access accelerates evidence gathering without allowing an uncertain hypothesis to alter production, and remediation remains subject to established approval and rollback controls. The decisive constraint is the high impact of an incorrect command during incomplete diagnosis. The principle is to separate reasoning and diagnosis from high-impact execution. This matters in production because incident urgency should not bypass deterministic safeguards that limit additional damage. Why the other options are wrong: - Option B: Command explanations improve transparency, but they do not prevent an incorrect hypothesis from causing damage through unrestricted access. - Option C: Rollback can be an appropriate remediation, but automatically assuming the latest deployment is causal is unsafe when evidence remains incomplete. - Option D: Logging supports later investigation, but it is a compensating control and cannot prevent harmful configuration changes during the outage.

Sample Question 6 — Developer Productivity & Operational Enablement

A travel platform has an intermittent booking failure spanning six services. About 0.4% of requests fail, primarily after a recent configuration rollout, but individual service dashboards remain within normal ranges. Raw logs total 18 TB per day and cannot fit into a useful model context. Logs and traces include request IDs, deployment versions, and timestamps. The operations team needs a defensible root cause, not merely a plausible summary. Which troubleshooting design is BEST?

  1. A. Summarize each service's daily logs independently, provide the six summaries to Claude, and select the explanation mentioned most frequently.
  2. B. Increase the model context allocation, sample equal volumes from every service, and ask Claude to infer the most likely failure sequence.
  3. C. Create a permanent vector index over all raw logs, retrieve semantically similar messages, and use similarity frequency as causal evidence.
  4. D. Correlate failed requests by trace identifiers and versions, retrieve evidence iteratively for explicit hypotheses, and preserve provenance and contradictions. (Correct answer)

Correct answer: D

Explanation: Correct answer (D): Trace-based correlation reconstructs individual cross-service failures, while hypothesis-driven retrieval limits context without losing request relationships, deployment versions, or contradictory evidence. The governing constraints are the intermittent cross-service symptom, massive log volume, and need for a defensible causal conclusion. The architectural principle is evidence-driven troubleshooting with correlation and provenance rather than indiscriminate context expansion. In production, this makes conclusions testable and avoids confusing frequent or semantically similar messages with causation. Why the other options are wrong: - Option A: Per-service summaries reduce context volume, but independent aggregation can erase the request-level sequence and rare correlations needed to explain a 0.4% failure. - Option B: Larger context and balanced sampling may expose more evidence, but random volume does not preserve causal request relationships and remains inefficient at this scale. - Option C: Semantic retrieval can find related symptoms, but similarity frequency does not establish event order or causality and may miss exact identifiers and versions.

Sample Question 7 — Developer Productivity & Operational Enablement

An enterprise has 70 repositories across mobile, data, embedded, and cloud teams. All teams must follow common security and release controls, but their build systems and domain procedures differ. A pilot used one universal instruction prompt; it has grown to 38 pages, contains conflicting commands, and consumes context even when most sections are irrelevant. Which enablement architecture should replace it?

  1. A. Maintain separate complete prompts for every repository and have each team manually copy enterprise controls into its local version.
  2. B. Shorten the universal prompt by removing detailed procedures and depend on developers to supply missing team instructions during each session.
  3. C. Create governed reusable components for enterprise controls, load owned domain modules when relevant, and permit reviewed repository-level extensions. (Correct answer)
  4. D. Place all standards in a semantic index and let Claude retrieve any instructions that appear related to the developer's current request.

Correct answer: C

Explanation: Correct answer (C): Modular, governed components preserve mandatory enterprise controls while loading specialized procedures only where applicable and allowing reviewed local requirements. The explicit constraint is consistency across heterogeneous repositories without conflicting or oversized context. The architectural principle is scoped reuse with clear ownership, versioning, and controlled extension. In production, modular configuration reduces duplication and context dilution while enabling enterprise changes to propagate through an auditable mechanism. Why the other options are wrong: - Option A: Repository-specific prompts fit local workflows, but copying enterprise controls creates duplication, inconsistent updates, and governance drift. - Option B: A shorter prompt reduces context use, but relying on developers to supply required procedures makes behavior inconsistent and weakens onboarding. - Option D: Retrieval can surface relevant guidance, but probabilistic retrieval alone may omit mandatory security controls that must be applied deterministically.

Sample Question 8 — Developer Productivity & Operational Enablement

A retailer piloted Claude-assisted development with 600 engineers. Organization-wide suggestion acceptance is 61%, and reported coding time fell 14%. However, two payments teams report longer reviews, and security rework increased after generated changes. Leadership wants an immediate rollout to 4,000 engineers. Telemetry cannot retain source code, secrets, or raw prompts. What should the architect recommend FIRST?

  1. A. Proceed with the rollout because aggregate acceptance and coding-time metrics demonstrate net productivity, then correct team-specific issues afterward.
  2. B. Pause all usage until the platform can retain complete prompts and generated source so every accepted suggestion can be reconstructed.
  3. C. Replace the current model with a more capable model across all teams and compare organization-wide acceptance after one month.
  4. D. Define net-value and safety measures, segment results by team and task risk, instrument privacy-preserving telemetry, and expand through gated phases. (Correct answer)

Correct answer: D

Explanation: Correct answer (D): A segmented, phased evaluation determines whether coding-time gains survive review, rework, defect, and security costs without collecting prohibited content. The decisive constraints are hidden adverse outcomes in payments teams, the proposed scale increase, and privacy limits on telemetry. The architectural principle is to establish multidimensional, representative evidence and regression gates before broad optimization or rollout. This matters because favorable averages can conceal high-risk segments whose remediation costs erase apparent productivity gains. Why the other options are wrong: - Option A: Fast expansion could capture aggregate gains, but the averages conceal increased security rework and review time in consequential payment workflows. - Option B: Complete reconstruction could aid diagnosis, but retaining raw prompts and source violates the telemetry constraint and is unnecessary for useful evaluation. - Option C: A stronger model might improve some tasks, but no evidence yet isolates model capability as the cause, and aggregate acceptance would still hide weak segments.

Keep Practicing

Take the 10-question CCAR-P quick-start test across all 7 domains, or return to the CCAR-P practice test hub for 200+ questions, exam details, and a study plan. Drill the other domains: Integration, Solution Design & Architecture, Evaluation, Testing & Optimization, Governance, Safety & Risk Management, Stakeholder Communication & Lifecycle Management, Claude Models, Prompting & Context Engineering, and Developer Productivity & Operational Enablement.