salesforce agentforce Practice Questions & Answers (Set 13) | CodeWme
#1 What is the purpose of the 'Agentforce Testing Center'? Select 1
✅ Answer: To batch test utterances against expected topics and actions to verify accuracy.
The Testing Center allows specialists to run batch tests of utterances to see if the correct topics and actions are triggered, ensuring regression testing reliability.
#2 When debugging an Agent interaction that provided an incorrect answer, where should the specialist look to see the internal 'Chain of Thought'? Select 1
✅ Answer: Conversation History / Event Logs
The Conversation History (and associated Event Logs) in the Agent Builder provides the 'Dynamic Debugging' view showing the plan, reasoning, and action outputs.
#3 Which Prompt Builder template type allows you to generate content based on multiple inputs, unrelated to a specific single record? Select 1
✅ Answer: Flex Template
Flex Templates are the most versatile, accepting multiple inputs and defined resources without being tied to a specific page context like Field Generation.
#4 True or False: Standard Agent Actions (like 'Identify Record') require you to manually create a Flow. Select 1
✅ Answer: False
Standard Actions are out-of-the-box capabilities provided by Salesforce that do not require manual Flow creation.
#5 Which element acts as the 'container' for a group of related actions and instructions that an Agent can perform? Select 1
✅ Answer: Topic
A Topic is the container that groups specific business capabilities (Actions) and the rules for using them (Instructions).
#6 Universal Containers (UC) stores case details and updates in several custom objects and would like its Agentforce Service Agent to be able to provide information based on them to its customers. Which best practice should UC follow to grant access to this information? Select 1
✅ Answer: Create a new permission set with the necessary object permissions and assign it to the Agentforce Service Agent user.
Standard permission sets for Agentforce are often locked or should not be modified. The best practice is to create a new, additive permission set with the specific custom object access required and assign it to the Agentforce user.
#7 When is the Agent-to-Agent (A2A) protocol an appropriate communication choice? Select 1
✅ Answer: When agents need to collaborate
The Agent-to-Agent (A2A) protocol is designed for delegation and collaboration between agents. Tool access and API invocations are typically handled by the Model Context Protocol (MCP).
#8 Universal Containers (UC) is setting up a new Agentforce Service Agent. The company has sensitive data stored internally and wants to ensure the agent cannot access it. What should UC do? Select 1
✅ Answer: Follow the principle of least privilege and avoid granting permissions to sensitive fields.
To prevent an agent from accessing sensitive data, you must restrict access at the Field-Level Security (FLS) level for the user the agent operates as.
#9 Cloud Kicks (CK) recently finished the development of a new prompt template. After deploying a prompt template from a sandbox to a production org, CK is getting an error related to the LLM used in the prompt template. What is the cause of the error? Select 1
✅ Answer: The name of the LLM does not match in sandbox and production.
The Prompt Template metadata references the API Name of the LLM. If the LLM configuration in Production has a different API Name than in Sandbox, the deployment will fail or the template will error.