> ## Documentation Index
> Fetch the complete documentation index at: https://docs.trusapien.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge Grounding

> Your Agents search your connected data sources for answers grounded in your organization's actual documents and information

Every result your Agents deliver can be grounded in what your organization actually knows. When Knowledge Grounding is active, your Agents search across your connected Connectors to find relevant documents, tickets, wiki pages, and records. The difference: answers based on your organization's actual data, not generic responses.

## How Knowledge Grounding works

Toggle Knowledge Grounding on from the chat input bar (the two-plug icon). A labeled pill appears confirming it is active. Your Agent searches across all connected data sources for information relevant to your request and weaves relevant findings into its response.

You do not need to tell your Agent which documents to check. Your Agent finds what it needs automatically.

## Filtering by source

When Knowledge Grounding is active, a filter panel appears with per-Connector toggles. Each connected source (Confluence, Jira, File, and others) appears as an individual toggle:

* **Blue toggles** are active: your Agent searches that source.
* **Grey toggles** are inactive: your Agent skips that source.
* **"Disable All"** turns off all sources at once.

This gives you precise control. Searching only your Confluence documentation? Toggle off everything else. Need information from Jira tickets specifically? Toggle Jira on and others off.

## Two data sources

Knowledge Grounding draws on two types of sources:

* **Connectors** link your existing tools: Confluence, Google Drive, Jira, Salesforce, and [39 total](/connectors/overview). Once connected, your Agents search indexed data from these tools automatically.
* **Vaults** hold documents you upload directly into Trusapien. Upload policies, reference materials, templates, or any documents your Agents should know about.

<Note>
  Knowledge Grounding searches data from your connected Connectors and Vault contents. It operates separately from files attached directly to a conversation via the paperclip icon. For large datasets, use the File Connector for indexed, searchable access.
</Note>

<Tip>
  Start by connecting your most-used tools (Confluence, Google Drive, Jira) and expand from there. The more relevant data sources you connect, the more grounded your results become.
</Tip>

<CardGroup cols={2}>
  <Card title="Connectors" icon="plug" href="/connectors/overview">
    Connect your existing tools and data sources
  </Card>

  <Card title="Vaults" icon="box-archive" href="/overview/platform/vaults">
    Upload and organize documents directly
  </Card>
</CardGroup>