Ridgeline's Data Library contains discharge instructions, clinic policies, and orthopedic specialty referral guidelines all together. An admin wants an agent answering general wellness questions to avoid retrieving orthopedic-specialty content by mistake. What should the admin configure?
Select an answer to reveal the explanation.
Short Explanation
Rather than nuking content another agent still needs, just point the wellness agent's retriever at the right shelf — scoping is what keeps different agents from tripping over each other's content.
Full Explanation
Retrievers can be scoped or filtered to specific sources or content types, which lets an admin ensure the wellness agent only pulls from the wellness-relevant material even though the specialty referral hospital's agent might still need the orthopedic guidelines from that same shared library. Deleting the orthopedic content outright would break grounding for any other agent, such as the specialty referral agent, that legitimately needs it — an overly destructive fix for what is really a scoping problem. Renaming documents has no defined effect on retriever behavior; retrieval relies on indexing and configured scope, not on filenames or titles being informally descriptive. Switching search type between keyword and vector changes how matching works, not which sources are eligible to be searched at all — search type and source scope are two independent configuration dimensions. The exam-relevant point is that multiple agents can share one Data Library while each retriever is independently scoped to the subset of content appropriate for that agent's purpose.