Free CCAR-F Exam Braindumps
Claude Certified Architect - Foundations Questions 26–30 (page: 7)

Updated On: 17-Nov-2026

Your scheduling agent uses get_available_slots(date, provider_id) to retrieve open appointment times, then book_appointment(provider_id, slot_time, patient_id) to reserve a slot. tickets show that 15% of booking attempts fall with "slot no longer available" because another user booked the slot between the availability check and the booking call. How should you r these tools?

  1. Modify book_appointment to return detailed failure information including currently available alternative slots when the requested slot is unavailable, enabling the agent to retry with a di time.
  2. Keep both tools but add retry logic to the agent's system prompt, instructing it to call get_available_slots again and select a different time if booking fails.
  3. Add a hold_slot(provider_id, slot_time) tool that creates a 60 second temporary reservation, requiring the agent to call it between checking availability and booking.
  4. Combine both tools into a single find_and_book_appointment that atomically checks availability and books, returning either the confirmed booking or available alternatives.

Answer(s): D

Explanation:

A: Modify book_appointment to return detailed failure information including currently available alternative slots when the requested slot is unavailable, enabling the agent to retry with a different time. Incorrect.
This improves recovery but does not fix the race condition between availability check and booking.
B: Keep both tools but add retry logic to the agent's system prompt, instructing it to call get_available_slots again and select a different time if booking fails. Incorrect.
This still suffers from the same race condition and relies on agent behavior rather than fixing the underlying issue.
C: Add a hold_slot(provider_id, slot_time) tool that creates a 60 second temporary reservation, requiring the agent to call it between checking availability and booking. Incorrect.
This reduces the issue but introduces additional complexity and still requires multiple steps that can fail.
D: Combine both tools into a single find_and_book_appointment that atomically checks availability and books, returning either the confirmed booking or available alternatives. Correct.
This eliminates the race condition by making the operation atomic , ensuring consistency and reliability.



Your agent has a log_workout tool that accepts exercise_type (string), value (number), and measurement (string). Production monitoring shows the agent frequently passes mismatched combinations-using measurement: "reps" for cardio exercises like running, or measurement: "miles" for strength exercises like bench press. Your exercises naturally divide into two categories: cardio (measured in time or distance) and strength (measured in reps and sets). 23% of tool calls have invalid combinations.
What approach would most effectively reduce these errors?

  1. Implement server-side validation returning descriptive errors for invalid combinations, allowing the agent to retry with corrections.
  2. Add enum constraints on measurement limiting values to "minutes", "miles", "reps", or "sets" to prevent arbitrary measurement strings.
  3. Add explicit examples to the tool description showing valid combinations (e.g., "For running: use minutes or miles. For push-ups: use reps") with constraints for each exercise category.
  4. Split into log_cardio_workout (with duration_minutes or distance_miles parameters) and log_strength_workout (with reps and sets parameters).

Answer(s): D

Explanation:

A: Implement server-side validation returning descriptive errors for invalid combinations, allowing the agent to retry with corrections. Incorrect.
This catches errors after they occur but does not prevent them, leading to wasted turns and inefficiency.
B: Add enum constraints on measurement limiting values to "minutes", "miles", "reps", or "sets" to prevent arbitrary measurement strings. Incorrect.
This restricts values but does not prevent invalid combinations (e.g., still allows "miles" for bench press).
C: Add explicit examples to the tool description showing valid combinations (e.g., "For running: use minutes or miles. For push-ups: use reps") with constraints for each exercise category. Incorrect.
Helpful guidance, but not enforceable—agents can still make mistakes.
D: Split into log_cardio_workout (with duration_minutes or distance_miles parameters) and log_strength_workout (with reps and sets parameters). Correct.
This enforces correctness at the schema level by eliminating invalid parameter combinations entirely , significantly reducing errors.



Your MCP server includes archive_file(file_id) and delete_file(file_id) tools. Production logs show the agent calls delete_file when users ask to "remove old backups," policy requires archiving backup files. Both tools currently have minimal descriptions: "Archives a file" and "Deletes a file." Which change most directly improves tool selection?

  1. Add a confirmation step that requires users to type "CONFIRM DELETE" before delete_file executes.
  2. Implement server-side validation that rejects delete_file calls for files tagged as backups, returning an error message suggesting archive_file.
  3. Expand tool descriptions to clarify use cases, adding guidance like "Do not use for backup files" to delete_file.
  4. Add few-shot examples to the system prompt demonstrating that requests involving "backup" or "old" should use archive_file.

Answer(s): C

Explanation:

A: Add a confirmation step that requires users to type "CONFIRM DELETE" before delete_file executes. Incorrect.
This prevents accidental execution but does not improve the agent’s tool selection decision .
B: Implement server-side validation that rejects delete_file calls for files tagged as backups, returning an error message suggesting archive_file. Incorrect.
This enforces policy after the wrong choice is made but does not directly improve initial selection.
C: Expand tool descriptions to clarify use cases, adding guidance like "Do not use for backup files" to delete_file. Correct.
Clear, specific descriptions directly influence the agent’s tool selection reasoning , making it less likely to choose the wrong tool.
D: Add few-shot examples to the system prompt demonstrating that requests involving "backup" or "old" should use archive_file. Incorrect.
Helpful, but less direct and less reliable than improving the tool descriptions themselves.



Your CRM agent's delete_contact tool handles requests like "delete the duplicate entry for Acme Corp." The database contains similarly named records (e.g., "Acme Corp," "Acme Corporation," "ACME Corp Inc."), and analytics show 8% of deletions are reversed within 24 hours due to misidentified records. Users have also complained that the current multi-step confirmation flow adds too much friction to routine cleanup tasks.
Which approach most effectively reduces the error rate while maintaining workflow efficiency?

  1. Present matched records with differentiating fields and require single-click confirmation of the intended target before executing deletion.
  2. Require users to supply the exact record ID from the CRM Interface rather than using natural language references to contact names.
  3. Deploy automated duplicate detection that identifies and merges probable duplicates, removing the need for manual deletion requests.
  4. Implement soft-delete with a 30-day recovery window so users can undo mistakes without slowing down the deletion workflow.

Answer(s): A

Explanation:

A: Present matched records with differentiating fields and require single-click confirmation of the intended target before executing deletion. Correct.
This directly addresses ambiguity by showing clear distinctions between similar records while keeping the workflow fast with a lightweight confirmation step.
B: Require users to supply the exact record ID from the CRM interface rather than using natural language references to contact names. Incorrect.
This reduces errors but adds significant friction and hurts usability for routine tasks.
C: Deploy automated duplicate detection that identifies and merges probable duplicates, removing the need for manual deletion requests. Incorrect.
Helpful as a separate improvement, but it doesn’t solve incorrect deletions during manual requests .
D: Implement soft-delete with a 30-day recovery window so users can undo mistakes without slowing down the deletion workflow. Incorrect.
This mitigates impact after errors occur but does not reduce the error rate itself .



After Implementing tool use with strict schema definitions, JSON syntax errors are eliminated, but 5% of extractions still have valid JSON with empty arrays or null values for required fields like citations and methodology. Spot-checking reveals that source documents contain this information, but in varied formats—Inline citations vs. bibliographies, methodology sections vs. details embedded in Introductions.
What's the most effective way to address these failures?

  1. Modify your schema to make citations and methodology optional, and flag Incomplete records for manual review rather than falling validation.
  2. Build a regex-based post-processing layer that scans source documents for citation patterns and methodology keywords, populating empty fields when the model falls to extract.
  3. Add few-shot examples demonstrating extractions from documents with varied structures—showing how to identify citations in different formats and locate methodology details across section types.
  4. Implement retry logic that re-sends requests when validation detects empty required fields.

Answer(s): C

Explanation:

A: Modify your schema to make citations and methodology optional, and flag incomplete records for manual review rather than failing validation. Incorrect.
This lowers data quality standards and avoids solving the extraction problem.
B: Build a regex-based post-processing layer that scans source documents for citation patterns and methodology keywords, populating empty fields when the model fails to extract. Incorrect.
Regex approaches are brittle and unreliable across varied formats, especially for complex structures like methodology.
C: Add few-shot examples demonstrating extractions from documents with varied structures—showing how to identify citations in different formats and locate methodology details across section types. Correct.
This directly improves the model’s ability to generalize across diverse document formats , addressing the root cause of missed extractions.
D: Implement retry logic that re-sends requests when validation detects empty required fields. Incorrect.
Retries without improving guidance will likely produce the same incomplete outputs.



Viewing page 7 of 31
Viewing questions 31 - 35 out of 149 questions


Post your Comments and Discuss Anthropic CCAR-F exam prep with other Community members:

AI Tutor AI Tutor 👋 I’m here to help!