A customer raises three separate issues during one session: a refund inquiry (turns 1-15), a subscription question (turns 16-30), and a payment method update (turns 31-45). At turn 48, the customer asks "What happened with my refund?" The conversation is approaching context limits. What strategy best maintains the agent's ability to address all issues throughout the session?
Answer(s): B
A: Implement sliding window context that retains the most recent 30 turns. Incorrect.This would likely drop the earlier refund discussion entirely, preventing accurate follow-up handling.B: Extract and persist structured issue data (order IDs, amounts, statuses) into a separate context layer. Correct.Structured persistence preserves critical state across multiple concurrent issues while minimizing context usage and enabling reliable recall later in the session.C: Rely on MCP tools to re-fetch relevant information on demand when the customer references earlier issues. Incorrect.Tools can retrieve system data, but not the conversational context, decisions, or prior explanations from earlier turns.D: Summarize earlier turns into a narrative description, preserving full message history only for the active issue. Incorrect.Narrative summaries may omit important structured details needed to accurately resume earlier issues.
Your update_user_profile tool accepts a user_id (required) and an optional fields_to_update object. In testing, Claude frequently omits user_id or passes incorrectly structured data. What is most critical for helping Claude understand what parameter values to provide?
Answer(s): A
A: Clear parameter descriptions explaining expected format, such as "user_id: UUID of the user to update (required)" Correct.Clear, human-readable descriptions are the most important guidance for helping Claude understand what values to provide and how parameters should be structured.B: Verbose parameter names encoding format hints, such as user_id_string_uuid_format Incorrect.Overly verbose names reduce readability and are less effective than proper descriptions.C: Strict JSON Schema type constraints marking user_id as required and defining fields_to_update as an object type Incorrect.Schema constraints help validation, but they don’t sufficiently explain semantic expectations like the required UUID format.D: Detailed error responses explaining why invalid parameter values were rejected Incorrect.Helpful after failure, but not the most critical factor for preventing mistakes initially.
Production monitoring shows your search_catalog tool fails 12% of the time: 8% are network timeouts that succeed when immediately retried, while 4% are query syntax errors from malformed user-provided filters that never succeed regardless of retry attempts. Currently, both error types are returned to the agent identically, causing it to waste turns retrying syntax errors and telling users to "try again later" for timeouts. How should you modify the tool's error handling?
Answer(s): C
A: Apply exponential backoff retry logic to all errors uniformly, returning a generic "service temporarily unavailable" message after max retries are exhausted. Incorrect.This wastes retries on non-recoverable syntax errors and hides useful diagnostic information from the agent.B: Return all errors with a retryable boolean flag and error type details. Incorrect.This improves clarity, but still pushes retry logic to the agent, which leads to inconsistent handling and wasted turns.C: Implement automatic retry with backoff for network timeouts inside the tool; return syntax errors immediately with parameter validation details. Correct.This cleanly separates responsibilities:Tool handles transient failures (timeouts) Tool immediately exposes non-retryable syntax errors with actionable feedback This prevents wasted retries and improves user-facing behavior.D: Add few-shot examples to your system prompt demonstrating how to distinguish network errors from syntax errors and handle each case appropriately. Incorrect.Prompt-based reasoning is unreliable for error handling compared to explicit tool-level logic separation .
Users frequently send ambiguous requests like "book a venue for the party" without specifying date, guest count, or budget. Your evaluation shows the assistant asks an average of 4.2 clarifying questions before taking any action, causing 35% of users to abandon mid-conversation. However, when you reduce questions, users sometimes receive recommendations that don't match their preferences. What's the most effective approach to improve this trade-off?
Answer(s): D
A: Instruct the assistant to state explicit assumptions based on conversation history, proceed with recommendations while inviting corrections, and reserve clarifying questions only for irreversible actions like confirming bookings. Incorrect.This improves transparency, but it still relies on assumptions that may not fully align with user intent, leading to partial mismatches.B: Configure the assistant to proceed with reasonable defaults (medium-sized venue, next weekend, moderate budget) without explicitly stating these assumptions, allowing users to provide corrections if results don't match expectations. Incorrect.Hidden defaults can reduce trust and lead to user confusion when outputs don’t match expectations.C: Implement a structured intake form that collects all required parameters (date, guest count, budget, venue type) upfront before the assistant begins providing any recommendations. Incorrect.This ensures accuracy but introduces too much friction early in the flow, increasing drop-off.D: Configure the assistant to consolidate all clarifying questions into a single compound question (e.g., "What date, guest count, and budget are you considering?") to reduce the total number of conversational turns. Correct.This reduces user fatigue while still explicitly capturing required constraints, improving both completion rates and recommendation accuracy without over-questioning.
Your document extraction tool uses ML models to extract invoice fields (vendor, amount, date). The models return confidence scores (0.0-1.0) for each extracted field. In production, you observe: (1) the agent proceeds with low-confidence extractions that are incorrect 23% of the time, and (2) the agent requests unnecessary human review for 31% of extractions that were actually correct. How should you restructure the tool's output?
A: Return fields with their raw confidence scores and add detailed few-shot examples to your system prompt demonstrating how to interpret different confidence ranges and when to request human review. Incorrect.This still leaves the decision logic ambiguous and pushes critical operational behavior into prompt interpretation, which is inconsistent and hard to control in production.B: Compute an aggregate extraction quality score across all fields and return it alongside the extracted values. Include a text summary describing the overall extraction reliability. Incorrect.A single aggregate score hides per-field uncertainty, which is exactly what is needed to avoid both false approvals and unnecessary reviews.C: Return fields with confidence scores, plus a requires_review boolean computed using your tested confidence thresholds, along with a review_reasons array explaining which fields triggered review. Correct.This directly encodes decision logic into structured output, ensuring consistent behavior while still preserving explainability at the field level.D: Return fields organized into verified and needs_verification objects based on confidence thresholds. Incorrect.This improves structure, but it hides the underlying confidence signals and reduces flexibility for downstream systems that may need finer-grained control or auditability.
Your agent includes an update_game_score tool that accepts game_date (string), home_team (string), and away_team (string) parameters. Production logs reveal recurring issues: the agent uses team nicknames instead of official names, applies inconsistent date formats, and selects the wrong game when teams have rematches in the same season. What tool interface change would effectively prevent these errors?
A: Add a season parameter to disambiguate rematches, and add a confirm_before_update flag that returns the resolved game details for the agent to verify before the score is committed. Incorrect.This improves safety via confirmation, but it still leaves ambiguity in identifying the correct game as a core issue.B: Add detailed examples to the tool description showing the required date format and complete list of official team names. Incorrect.Examples help, but they don’t reliably prevent structural errors like wrong team identity or ambiguous match selection.C: Add enum constraints listing valid team names for both team parameters, and add a regex pattern enforcing ISO 8601 format for the date parameter. Incorrect.This improves input validation, but it still does not solve the deeper problem of disambiguating multiple games between the same teams .D: Replace the three parameters with a single game_id parameter and a separate search_games lookup tool that returns matching game IDs. Correct.This removes ambiguity entirely by forcing the agent to resolve the exact entity first, then operate on a unique identifier, eliminating nickname errors, date mismatches, and rematch confusion.
Your resource allocation tool returns a simple acknowledgment message after provisioning is requested. Users frequently approve allocations and immediately ask "how much did that cost?" or "which project was that?" -indicating they confirmed without understanding the request. What tool design change would most effectively address this?
A: Add a user_acknowledged: boolean parameter that must be set true, with instructions for the agent to only set it after the user explicitly confirms they reviewed the details Incorrect.This relies on agent discipline and is easy to bypass or mis-handle; it doesn’t guarantee users actually see or understand key details.B: Implement a 60-second hold before execution completes, allowing users time to review pending allocations and cancel if needed Incorrect.Delays alone don’t improve understanding; they only slow execution and may frustrate users without ensuring informed approval.C: Add a detail_level parameter with options "minimal" or "comprehensive" that controls how much context the agent presents in confirmations Incorrect.This improves formatting flexibility but does not ensure critical cost and impact information is consistently surfaced before approval.D: Return structured data including cost estimate, target project, resource specifications, and impact summary in the tool response Correct.This ensures that every allocation includes explicit, structured, reviewable details, reducing uninformed approvals and enabling users to understand cost and impact before proceeding.
Your conversation history includes two types of content: persistent story elements (character backgrounds, plot structure, world rules) that must remain consistent throughout, and extensive brainstorming discussion that's mostly ephemeral. After 40+ turns, you're hitting context limits and users report the assistant "forgets" established character traits, breaking narrative consistency. Which approach best ensures persistent story elements remain available to the model while reclaiming context space?
A: Separate persistent story elements into a retained "story bible" section at context start, applying trimming or summarization only to brainstorming discussion. Correct.This directly solves the problem by preserving high-value, invariant narrative constraints (characters, rules, canon) while allowing lower-value conversational content to be compressed. It maintains continuity without losing essential story logic.B: Store all history in a vector database and retrieve semantically similar passages for each new message, replacing conversation history with retrieved chunks. Incorrect.This risks missing globally important story rules because retrieval is similarity-based, not constraint-aware. Core canon details may not always be retrieved.C: Apply a sliding-window approach keeping only the most recent 25 turns, relying on the model to infer earlier context from recent discussion flow. Incorrect.This will almost certainly break long-range consistency, since key character and world rules will be dropped.D: Summarize the entire conversation history into a condensed synopsis every 20 turns, replacing the full history to free up tokens. Incorrect.General summarization tends to lose fine-grained canonical details, which are exactly what must remain stable for narrative consistency.
Post your Comments and Discuss Anthropic CCA-F exam prep with other Community members:
Our website is free, but we have to fight against AI bots and content theft. We're sorry for the inconvenience caused by these security measures. You can access the rest of the CCA-F content, but please register or login to continue.