CCDV-F Agents and the Agent SDK Practice Question
A developer is building a customer support agent using the Anthropic SDK. The agent must call a 'get_order_status' tool that requires an order ID. During testing, the agent sometimes invents plausible-looking order IDs instead of asking the user. Which change to the tool definition will most directly reduce this behavior?
⚠ Common exam trap
The trap here is assuming that changing a tool's name or schema shape will fix a behavioral problem that is actually caused by missing usage instructions in the description.
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
✓
Add a detailed 'description' field to the tool that explains when to use it and explicitly instructs the model not to guess missing parameters.
The tool description is the model's instruction manual for when and how to invoke a tool. By explicitly stating that the agent must request an order ID from the user rather than guessing, the developer directly targets the observed hallucination. Schema changes, token limits, and renaming do not convey that behavioral rule.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase the 'max_tokens' setting so the model has more room to reason before emitting the tool_use block.
Why it's wrong here
Max tokens controls the total length of the generated response, not the quality of parameter reasoning. The agent is not failing because it ran out of space; it is failing because it lacks guidance on missing data. Adding more token budget does not supply that instruction and may even give the model more room to produce a fabricated order ID.
- ✗
Set 'input_schema' to an empty object so the model is forced to ask the user for all parameters.
Why it's wrong here
An empty input schema removes the model's understanding of what parameters the tool accepts, making correct invocation impossible. The model would not know that an order ID is required or what format it should take. This does not prevent hallucination; it simply removes the structure needed for a valid call and would likely cause malformed tool_use blocks.
- ✓
Add a detailed 'description' field to the tool that explains when to use it and explicitly instructs the model not to guess missing parameters.
Why this is correct
The tool description is the primary place to encode usage guidance, including when the tool is appropriate and how to handle missing required inputs. Stating that the model must not fabricate an order ID and should instead ask the user directly addresses the observed failure mode without altering the schema or adding external logic.
- ✗
Change the tool name to 'lookup_order' so it sounds more like a retrieval operation.
Why it's wrong here
Tool names are identifiers, not behavioral instructions. Renaming the tool does not tell the model what to do when a required argument is absent, so the underlying hallucination pattern remains. While clear naming is good practice, it is far less effective than explicit description text that defines the missing-parameter policy.
About these practice questions
This CCDV-F question is part of Courseiva's 257-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 CCDV-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 CCDV-F exam.