CCAR-P Hard Questions — 25 Difficult Questions

Question 1

An architect needs to build an agent that handles complex, multi-step data migrations where each step depends on the output of the previous one. Which approach is most robust for ensuring the agent doesn't lose track of the long-term goal during execution?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Plan-and-Execute Pattern. For multi-dependency tasks, a Plan-and-Execute pattern is superior because it separates the high-level strategy from the low-level tool interactions. By maintaining an explicit plan that is updated after each step, the agent can track its progress against the original goal and adjust its future actions based on the results of completed tasks.

Answer

Reactive Tool Calling (ReAct)

Answer

Zero-shot prompting with all tools

Answer

Few-shot prompting with migration examples

Answer

Parallel Sub-tasking

Question 2

Refer to the exhibit. An application frequently hits rate limits during peak hours. What is the most robust way to improve operational reliability?

Accepted Answer

Implement exponential backoff with jitter in the API client layer.. Handling rate limits through exponential backoff is a standard practice to maintain system stability. When the API returns a rate limit error, the client should wait for the specified duration or use a backoff strategy before retrying. This approach prevents overwhelming the service, respects API quotas, and ensures that the application recovers gracefully from traffic spikes, ultimately leading to a more resilient and professional-grade production architecture.

Answer

Catch the error and retry the request immediately in a loop.

Answer

Increase the timeout duration for all API calls in the application.

Answer

Switch to a synchronous architecture to serialize all API requests.

Question 3

An enterprise is deploying Claude for high-stakes financial analysis. Which TWO governance controls should be implemented to mitigate the risk of model hallucinations and ensure factual accuracy?

Accepted Answer

Implement Retrieval-Augmented Generation (RAG) to ground responses in internal trusted knowledge bases.. Mitigating hallucination risk requires a multi-layered approach involving technical constraints and validation processes. Implementing robust Retrieval-Augmented Generation (RAG) grounds the model's responses in verified source documents, while secondary verification steps add a deterministic layer to the output. These controls are essential in financial services where incorrect data can lead to severe regulatory penalties, financial loss, and significant reputational damage to the organization.

Answer

Increase the temperature parameter to 1.5 to maximize response creativity.

Answer

Require human intervention for every prompt sent to the API to guarantee zero errors.

Answer

Disable all safety filters to allow the model to process complex financial jargon.

Question 4

Which THREE strategies should be employed to securely manage sensitive data within an agentic workflow?

Accepted Answer

Apply PII masking/redaction before passing text to the LLM.. Securing agents requires a layered approach: PII masking before sending to the model keeps data private; granular access control ensures the agent only touches data it is authorized to see; and audit logging provides traceability for every decision made. These components ensure that even if the agent is compromised or hallucinates, the impact is contained, data privacy is maintained, and there is a clear record of actions taken for compliance and forensics.

Answer

Enable the model to have full administrative access to the file system.

Answer

Embed all user data directly into the system prompt for faster access.

Question 5

An architect is defining the Shared Responsibility Model for a company deploying Claude via the Messages API. Which THREE tasks are the sole responsibility of the customer (the 'User') rather than Anthropic?

Accepted Answer

Classification and sanitization of PII within input prompts.. Understanding the Shared Responsibility Model is essential for risk management in AI deployments. While Anthropic secures the base model and underlying infrastructure, the customer remains responsible for the data they input, how they configure access to the API, and the specific ways the model's output is integrated into their business processes.

Answer

Physical security of the data centers housing the TPU/GPU clusters.

Answer

Fine-tuning the base model's Constitutional AI principles.

Question 6

To ensure long-term maintainability and performance of LLM-based applications, which THREE architectural patterns should architects recommend? (Select THREE)

Accepted Answer

Decouple prompt management and model selection from core application logic.. Decoupling model logic, implementing robust caching, and establishing monitoring are vital for long-term sustainability. Decoupling allows for model upgrades without massive refactoring, caching reduces latency and costs for repetitive queries, and monitoring ensures that performance degradation is caught immediately. These patterns transform 'experimental' LLM features into production-grade systems that developers can manage efficiently, reducing the technical debt typically associated with quickly-built AI integrations.

Answer

Hardcode system prompts to ensure the model behavior cannot change over time.

Answer

Use the largest available context window for every single request to maximize intelligence.

Question 7

Refer to the exhibit. The application is hitting rate limits during peak hours. What is the best architectural change to improve operational resilience?

Accepted Answer

Implement an exponential backoff strategy with jitter.. Implementing an exponential backoff strategy with jitter is the industry-standard approach for handling 429 rate-limiting errors. Unlike static retries, this approach prevents the 'thundering herd' problem, where multiple failed requests attempt to reconnect simultaneously, further stressing the service. This enhances operational stability, ensures more graceful handling of high-traffic scenarios, and improves the overall reliability of the system, which is a key requirement for professional architects.

Answer

Increase the timeout duration of the API calls indefinitely.

Answer

Bypass the API gateway and send requests directly to the model's backend IP.

Answer

Disable all retries to prevent the application from crashing.

Question 8

Your team is deploying an application that uses PII in prompts. A stakeholder asks how you are mitigating the risk of data leakage. How do you respond?

Accepted Answer

Explain the use of PII masking and data sanitization pipelines.. The correct response involves explaining a combination of data sanitization, prompt masking, and secure infrastructure design. This is critical because PII handling is a high-liability area. By detailing a defense-in-depth strategy, the architect provides the necessary assurance that privacy is treated with the highest priority, which is vital for maintaining organizational compliance and preventing severe legal or reputational damage during the application's lifecycle.

Answer

Assure them that the model is smart enough to handle PII.

Answer

Tell them the model does not store any data anyway.

Answer

Suggest moving the project to an on-premise LLM to ensure security.

Question 9

Refer to the exhibit. An agentic workflow encounters an error immediately after this message is generated by Claude. No further messages are sent to the API. What is the most likely cause of the failure in the orchestration logic?

Accepted Answer

The orchestrator failed to send a tool_result message. The Anthropic Messages API requires a specific sequence for tool use. After the assistant generates a 'tool_use' block, the orchestration layer must execute the tool and return a 'tool_result' message to the model before the assistant can continue its turn. Failing to provide this required response breaks the synchronous nature of the tool-calling conversation flow.

Answer

The input parameters do not match the JSON schema

Answer

The tool_use ID is missing the required prefix

Answer

The assistant content block lacks a stop_reason

Question 10

A team is designing a Claude-based assistant for internal developers. They want to ensure the assistant is reliable and safe in production. Which TWO practices should they implement? (Choose two.)

Accepted Answer

Implement automated evaluation and monitoring of model outputs for quality and safety.. Input validation and sanitization reduce prompt injection risks, while automated evaluation and monitoring catch quality and safety regressions. Together, they provide proactive defense and continuous oversight, which are essential for a reliable and safe production assistant. Other practices like high temperature or disabling logging would undermine these goals.

Answer

Disable all logging to protect user privacy.

Answer

Set the model's temperature to a high value to encourage diverse responses.

Answer

Log all user prompts and model outputs for debugging and auditing.

Question 11

An agent uses a retrieval tool that returns the top 20 chunks for any query. In production, the agent frequently cites irrelevant chunks and sometimes misses the correct answer even when it is present in the corpus. The corpus contains documents with overlapping terminology. Which architectural change most improves answer grounding without increasing the number of retrieved chunks?

Accepted Answer

Add a re-ranking stage that scores the retrieved chunks against the query and passes only the highest-scoring subset to the model.. The retrieval step is high-recall but low-precision, which is why irrelevant chunks are cited and correct ones are missed amid overlapping terminology. Adding a re-ranking stage that scores candidates against the query and passes only the top subset improves precision without retrieving more chunks, directly improving grounding and citation quality.

Answer

Embed the entire corpus into the system prompt so the model always has full context available.

Answer

Increase the retrieval count to 50 so the correct chunk is more likely to be included somewhere in the context.

Answer

Lower the model's temperature to zero so it sticks more closely to the retrieved text.

Question 12

You are architecting a Claude-based agent for a regulated financial client that performs long-running portfolio rebalancing workflows. A compliance requirement mandates that no single trade instruction may be executed unless it is cryptographically traceable to the exact model-generated intent that produced it. The agent uses the Messages API with tool use, and several downstream services consume tool calls asynchronously. Which architectural mechanism best satisfies this requirement while preserving agent autonomy?

Accepted Answer

Persist every assistant turn containing a tool_use block, its tool_use id, and the matching tool_result, forming an immutable audit chain keyed by the tool_use id.. Binding each tool call to the platform-generated tool_use id and persisting the surrounding assistant turn creates a deterministic chain from model intent to executed instruction. Because the id appears in both the tool_use block and the tool_result, downstream services and auditors can reconcile them without trusting model-authored text. Caching, summaries, or self-hashes do not establish that binding.

Answer

Require the model to emit a SHA-256 hash of its own reasoning text inside the tool input, and verify that hash in the downstream trade service before execution.

Answer

Route all trades through a single orchestrator agent that logs a natural-language summary of each decision to an append-only ledger after the trade is confirmed.

Answer

Enable prompt caching on the system prompt so that every trade instruction inherits a stable cache key that can be presented to auditors as proof of origin.

Question 13

A global bank is deploying Claude for internal document summarization. The risk committee requires controls to prevent unauthorized data leakage and ensure compliance with regional privacy laws. Which TWO controls are most effective for mitigating these risks? (Choose two.)

Accepted Answer

Configure the API to automatically redact personally identifiable information (PII) before sending data to Claude.. The most effective controls are DLP scanning and automatic PII redaction. DLP scanning inspects and blocks sensitive data in both directions, while redaction prevents sensitive data from ever reaching the model. Together, they provide strong preventive measures against data leakage and help meet regional privacy requirements. Other options are either misaligned with the deployment model or are detective/administrative rather than preventive.

Answer

Enable verbose logging of all prompts and responses for forensic analysis.

Answer

Require all users to sign an acceptable use policy before accessing the summarization tool.

Answer

Use a private, dedicated instance of Claude with no internet access and strict network segmentation.

Question 14

A retail company is building a governance program for a customer-facing Claude agent that can issue refunds and update order records. The risk committee wants controls that limit the blast radius of a compromised or misbehaving agent. (Choose two.)

Accepted Answer

Require human approval for refunds above a defined monetary threshold before the action is executed.. Blast-radius reduction requires preventive controls that bound what the agent can execute. Least-privilege tool scopes with per-action authorization shrink the callable action set, and threshold-based human approval stops high-value refunds before they happen. Logging, policy publication, and periodic transcript review are detective or administrative measures that document or discourage misuse but do not constrain the agent's real-time capabilities, so they fail the risk committee's objective.

Answer

Enable verbose debug logging of every prompt and response for later forensic review.

Answer

Schedule quarterly reviews of the agent's conversation transcripts to identify emerging misuse patterns.

Answer

Publish an acceptable use policy that prohibits employees from manipulating the agent.

Question 15

A fintech company's risk committee is operationalizing a governance program for Claude-powered customer support agents. They must demonstrate to regulators that model behavior changes are tracked, attributable, and reversible. Which TWO practices best satisfy this requirement? (Choose two.)

Accepted Answer

Run a fixed regression suite against each candidate model version and archive the scored results before promoting it.. Attributable, reversible model change management requires two things working together: a frozen, dated artifact in production and comparable evidence captured before promotion. Pinning to dated model identifiers supplies the frozen artifact and the rollback target, while a fixed regression suite run against each candidate supplies the before-and-after evidence. Together they let the risk committee answer what changed, who approved it, and how to undo it.

Answer

Route every request through a gateway that logs the full request and response payloads with the model identifier attached.

Answer

Enable prompt caching on all production requests so that previously validated prompts are reused verbatim.

Answer

Store the system prompt in a version-controlled repository and require pull-request review before it is merged.

Question 16

A multinational bank uses Claude via Amazon Bedrock with a cross-region inference profile so that requests from its Frankfurt branch may be served by any of several EU regions. The bank's regulator requires that no customer PII leave the European Economic Area and that every inference be attributable to a named human approver. The architecture team proposes Anthropic-side controls to satisfy both requirements. Which TWO approaches are technically valid for meeting these requirements? (Choose two.)

Accepted Answer

Deploy the application in a single AWS region, route all traffic there, and gate each request behind an approval workflow whose identity is written to the application audit log.. Meeting a data residency mandate requires controlling where inference executes, not merely filtering content, so regional scoping of the inference path or a single-region deployment is necessary. Meeting an attribution mandate requires a durable record linking each call to an identified human, which comes from audit logging of the invoking principal or an approval workflow. Content filters and prompt instructions influence what the model says but cannot prove where processing occurred or who authorized it.

Answer

Enable Anthropic's Zero Data Retention option in the Anthropic Console for the bank's organization.

Answer

Configure the Bedrock guardrail with a PII filter set to ANONYMIZE on the input and the model response.

Answer

Add a system prompt instructing Claude to refuse any request containing a customer name or account number.

Question 17

You are preparing a go-live readiness review for a Claude-powered advisory assistant used by field staff. The steering committee wants assurance that operational ownership is clear before launch. Which two artifacts are most essential to demonstrate that the lifecycle handover from project to operations is complete? (Choose two.)

Accepted Answer

A documented service ownership agreement naming the accountable team, support hours, and service level objectives.. A complete handover requires both a named owner and the operational procedures that owner will follow. The service ownership agreement establishes accountability, support hours, and service level objectives, while the signed runbook documents incident response, escalation, and maintenance. Together they give the steering committee confidence that the assistant will be sustained after the project team transitions away, rather than drifting without ownership or support coverage.

Answer

A list of all stakeholders who attended the project kickoff meeting.

Answer

A complete archive of every prompt version ever tested during development.

Answer

A slide deck summarizing the assistant's key features for the marketing team.

Question 18

Your organization is implementing an AI governance board to oversee Claude deployments. What is the most effective approach for managing prompt injection risks in production?

Accepted Answer

Implement a separate, lightweight model to inspect all user inputs for malicious intent.. Managing prompt injection requires an architecture that separates untrusted user input from system instructions. Implementing a strict input sanitization layer combined with a robust system prompt structure ensures that the model can distinguish between instructions and data. This defense-in-depth approach is vital for enterprise security, preventing malicious actors from overriding model behaviors and ensuring that safety guardrails remain effective against adversarial manipulation attempts.

Answer

Rely solely on the model's inherent ability to distinguish between user intent and instructions.

Answer

Force users to sign a legal agreement regarding acceptable use before accessing the UI.

Answer

Disable system prompts entirely to prevent attackers from overriding them.

Question 19

When designing a system for sensitive data analysis using Claude, which THREE architectural choices represent best practices for safety and risk management?

Accepted Answer

Use ephemeral compute environments to process requests.. Best practices for sensitive data analysis involve minimizing exposure, ensuring robust logging, and maintaining clear provenance. By using ephemeral environments, strict access controls, and comprehensive auditing, organizations can limit the risk of data leakage. These architectural decisions are foundational to maintaining a secure posture, ensuring that data is protected throughout the entire lifecycle of the request, from initial ingestion to final output generation and storage.

Answer

Store all raw user input in unencrypted public buckets.

Answer

Disable all authentication to improve system latency.

Question 20

Refer to the exhibit. A stakeholder is confused about this error. How do you explain this technical constraint to them?

Accepted Answer

Explain that the model has a fixed 'memory' limit and suggest summarizing the input content.. Explaining the context window requires relating a technical limit to the practical capacity of the system. By using the analogy of a 'working memory' limit, you help stakeholders visualize why the system cannot process unlimited information at once. This frames the error as a manageable constraint rather than a bug, allowing for discussions on summarization, document chunking, or architectural changes to handle larger datasets effectively without violating model capacity limits.

Answer

Blame the provider for not having enough memory for our specific application needs.

Answer

Suggest that we switch to a model with a larger context window, regardless of cost.

Answer

Tell the stakeholder that the code is wrong and needs to be rewritten entirely.