An Agentforce needs to create a Sales Email with a custom prompt template. They need to ground on the following data. Opportunity Products Events near the customer Tone and voice examples How should the Agentforce Specialist obtain related items?

To ground a prompt template with diverse and dynamic data such as related records (Opportunity Products), external information (Events), and specific examples (Tone), the most effective method is to use a Flow. A prompt-initiated autolaunched Flow can be configured to execute complex logic, including querying related Salesforce records, making callouts to external APIs to fetch event data, and retrieving text from other objects or metadata. The outputs of this Flow are then passed directly into the prompt template as grounding data, ensuring the generated Sales Email is contextual, accurate, and adheres to the desired tone.

B. Create a flex template that takes the records in question as inputs. A prompt template (whether "flex" or standard) is the consumer of the data, not the mechanism that obtains it. This option fails to describe how the data is fetched.
C. Utilize a standard email template and manually insert the required data fields. This describes a manual, non-AI process. It contradicts the question's premise of using a custom prompt template for AI-powered email generation and grounding. ---

1. Salesforce Help Documentation

"Ground Your Prompts with Data Using a Flow."

Section: "Ground Your Prompts with Data Using a Flow"

Content: This document explicitly states

"To ground a prompt template in a Flow

create a prompt-initiated autolaunched flow... Your flow can search for records

summarize data

run business logic

and more. The flow’s outputs are then fed into your prompt template." This directly validates that a Flow is the correct tool to fetch and provide the required data.

2. Salesforce Help Documentation

"Prompt Template Resources."

Section: "Resource Types"

Content: This page lists "Flow" as a primary resource type for prompt templates. It specifies

"Use a flow to run complex business logic and return the output to the prompt template." This confirms that using a Flow is the officially sanctioned method for complex data retrieval for grounding purposes

as described in the scenario.