Knowledge

Overview

The Knowledge section is where you manage the documents and memory that agents use as context when generating responses. Agents with RAG (Retrieval-Augmented Generation) enabled will search relevant knowledge before answering.

Documents Tab

Uploading Documents

1. Go to Knowledge in the sidebar.
2. Ensure the Documents tab is selected.
3. Click Upload Document and select a file from your machine.
4. The document is processed and indexed for retrieval.

Supported Formats

Common document formats (PDF, DOCX, TXT, etc.) are supported for upload.

Tagging Documents

Tags control which agents can use a document as context:

1. Click the tag icon on a document row to open the Edit Tags modal.
2. Add or remove tags.
3. Save.

In the Agent → Knowledge tab, assign the same tags to an agent so it retrieves documents with those tags during conversations.

Viewing a Document

Click on a document to open a detail view with its content and metadata.

Memory Tab

The Memory tab shows persistent memory entries stored for agents. Memory allows agents to recall facts across conversations without relying on uploaded documents.

Entries in memory are managed automatically by agents configured with memory enabled, or can be reviewed and cleared from this view.