Data Library source types
Key Takeaway
Each data library uses exactly one source type: Knowledge articles, uploaded files, or a custom retriever. You can't combine sources in one library, and you can't change a library's source later.
An Agentforce Data Library, or data library, connects AI features such as agents to your trusted content, so their answers are grounded in your data. The source type you build a library on controls what gets indexed and what you can configure.
What a data library does
A data library turns unstructured or semi-structured content, such as documents or long text fields, into searchable information for AI features.
When you save a library, Data 360 indexes its source and automatically creates a search index and a retriever for it. In an agent, the Answer Questions with Knowledge action uses the library's search index and retriever to answer the user's question with data from that library.
Every library type requires Data 360 to be set up. Data libraries also consume Data 360 credits.