Where in the Agentforce/Prompt Builder workflow is a Data 360 retriever actually connected so its results reach the model at run time?
Select an answer to reveal the explanation.
Short Explanation
A retriever doesn't do anything on its own — it has to be wired in as a grounding source on a specific prompt template or action before its results ever reach the model.
Full Explanation
A retriever is wired in as a grounding source for a specific prompt template, or referenced by an agent action that uses that template, so that at run time its retrieved chunks are inserted into the context sent to the model alongside the user's query and any other grounding. This is a deliberate, per-template configuration step, not a passive report attachment — reports and dashboards are for analytics on structured data, not for delivering retrieval results into a live conversational prompt. It is also not something that, once configured anywhere, silently applies org-wide to every prompt with no further reference; each prompt template that needs Data 360 grounding must explicitly incorporate the retriever, otherwise unrelated prompts would unpredictably pull in irrelevant context. Topic classification descriptions serve a different purpose entirely — they help the reasoning engine decide which topic a user's utterance belongs to, and are not where retrieval configuration lives. For the exam, remember that grounding a specific prompt or action is an explicit configuration choice tying a retriever to that particular piece of work.