You need to ensure that an agent's "Code Interpreter" environment has access to a specific Python library that is not included in the default sandbox. How should you handle this?
check_circle
Correct AnswerA
You can upload private or specific Python packages as file resources and associate them with the Code Interpreter tool.
Question 92
You are deploying an agent to a region that does not yet support the "Code Interpreter" tool. What is the most reliable fallback?
check_circle
Correct AnswerA
If a built-in tool is unavailable, you can replicate the logic (e.g., Python execution) in your own API and call it via Function Calling.
Question 93
Which status in the Agent Run lifecycle indicates that the system is trying to find available compute resources to start the run?
check_circle
Correct AnswerC
The 'queued' status means the run is waiting in line for the underlying infrastructure to begin the execution.
Question 94
You are building an agent that needs to create a summary of a 10-minute audio recording. Which workflow is most efficient?
check_circle
Correct AnswerB
Transcribing audio to text first is the standard pattern for enabling an LLM to reason over and summarize spoken content.
Question 95
You need to share an AI Search index across three different AI Foundry projects. Where should the Search service be linked?
check_circle
Correct AnswerB
Linking resources at the Hub level allows all projects within that Hub to share the same compute, search, and storage.
Question 96
Which Azure AI Search property must be set to true on a field to allow it to be used in a search.isenumerable or facet query?
check_circle
Correct AnswerB
The facetable property allows the search service to provide hit counts for categories (facets) based on the values in that field.
Question 97
Which status in the Agent Run lifecycle indicates that the system is waiting for the developer to provide the output of a tool?
check_circle
Correct AnswerD
requires_action' signals that the model has called a function and is waiting for the client to return the result.
Question 98
You are configuring a Content Filter for a healthcare app. You want to block all content related to "Medical Advice" generated by the AI. What should you use?
check_circle
Correct AnswerB
Custom Blocklists allow you to define specific phrases or topics that the model is prohibited from outputting.
Question 99
You are building an agent that needs to convert unstructured text into a specific JSON format for a downstream database. Which prompting technique is most effective?
check_circle
Correct AnswerD
Few-Shot Prompting (providing examples of the desired input/output pairs) is the best way to teach the model a specific output format like JSON.
Question 100
In the "Azure AI Agent Service", what is the primary benefit of using "Code Interpreter" for solving a complex physics equation?
check_circle
Correct AnswerB
LLMs can make arithmetic errors; Code Interpreter ensures 100% accuracy by executing actual code to solve the math.
