CCAR-F Agentic Architecture and Orchestration Practice Question
An agent tasked with data analysis repeatedly fails because the 'tool_result' it receives is too large for the context window, causing subsequent calls to truncate. What is the most effective architectural adjustment to resolve this?
⚠ Common exam trap
Candidates frequently assume that passing raw, large datasets directly back into the context window is safe, ignoring context length limitations and truncation risks.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Use the 'tool_result' to return a summary or a file path instead of the full raw data.
Large tool outputs can quickly consume the context window, leaving no room for the model's reasoning or future tool calls. Architects should implement a 'Summarize-or-Reference' pattern where the full data is stored externally and only a concise summary or a pointer is returned to the agent within the message history.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Switch to a model with a smaller context window to force more efficient coding.
Why it's wrong here
Switching to a smaller context window would exacerbate the problem rather than solving it. The agent would have even less room to process the large data, leading to more frequent failures and a complete inability to handle complex analysis tasks that require significant historical context.
- ✓
Use the 'tool_result' to return a summary or a file path instead of the full raw data.
Why this is correct
Returning a summary allows Claude to understand the key characteristics of the data without overwhelming its context window. If the model needs specific details, it can then use a different tool to query a specific subset of the data, maintaining a high signal-to-noise ratio in the conversation.
- ✗
Format the tool output as a series of multiple user messages to spread the load.
Why it's wrong here
Spreading the data across multiple messages does not reduce the total number of tokens in the context window. Claude's context is the sum of all messages in the history; therefore, this approach would still lead to truncation once the total token limit is reached, regardless of message count.
- ✗
Request that Claude only uses the last 100 lines of the tool output in its reasoning.
Why it's wrong here
Asking the model to ignore part of its context is unreliable and does not solve the underlying technical issue of the context window being full. The API will still charge for the tokens, and if the window is exceeded, the model simply won't receive the earlier parts of the conversation.
Visual reference
About these practice questions
This CCAR-F question is part of Courseiva's 271-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Anthropic exam blueprint
This CCAR-F practice question is part of Courseiva's free Anthropic certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the CCAR-F exam.