Data 360 Fundamentals20% of the exam

Agentforce Data Library setup and considerations

How an Agentforce Data Library turns Knowledge articles or uploaded files into a working RAG setup: what Salesforce provisions in Data 360, who builds a library and where, file limits, the locked source type, connecting it to an agent, and when to outgrow it with hand-built retrievers.

Official exam objective
Explain the considerations of Agentforce Data Library and its concepts.

~8 min read · 28 practice questions ·

What the exam expects

  • Explain what an Agentforce Data Library does and list what Salesforce provisions in Data 360 when you create one.
  • Create a library (who, where, which sources, file size limits) and know that its source type is locked.
  • Connect a library to an agent and explain why content is not answerable straight after upload.
  • Explain how retrieval from a library respects data access.
  • Decide when a library is enough and when to build retrievers in AI Models, or to use an agent action instead.

Key terms: grounding adds domain knowledge to a prompt so the LLM answers from your data; chunk is a small, self-contained passage of a document; search index stores the chunked, vectorized content; retriever searches an index and returns relevant results to AI features; data library is the quick-start bundle that builds all of these for you.

Key facts

Setup location
Detail
Setup, Agentforce Data Library, New Library; or the Knowledge tab in Agentforce Builder
Permissions to build
Detail
System Administrator plus Data Cloud admin (now Data Cloud Architect)
File size limits
Detail
4 MB for text or HTML, 100 MB for PDF
Source type
Detail
One source per library, can't be changed; each feature uses one library at a time
Web search
Detail
Not supported for new libraries; use the Search the Web standard action
Going live on an agent
Detail
Knowledge tab, select library, save, Activate
Access
Detail
Governance policies and row-level access checks apply to retrieved content
Sources for this section (7)
  • Why might an agent fail to find answers minutes after several hundred PDFs are uploaded?
  • When should a specialist build retrievers in AI Models instead of relying only on a library?
  • Which four components does a library provision automatically in Data 360?
  • Why is a data library the wrong fit for telling a customer their order's ship date?

The questions are free to read. Answers with their sources and self-grading are included with the pack.

Keep reading with the Agentforce Specialist pack

  • Core concepts
  • Exam traps
  • Worked scenario

The pack includes:

  • All 26 lessons of the Agentforce Specialist guide, with reading progress
  • Interactive decision walks and step-by-step sequence steppers
  • Inline practice questions and one-click topic practice (28 questions for this lesson)
  • An official source cited for every claim
  • An AI tutor that answers from the guide
  • Unlimited timed practice exams

Practice for the Agentforce Specialist exam