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

# Knowledge Base

> Give your agent product knowledge

Add content your agent can reference when answering questions.

## Sources

| Source              | Best for                                   | Updates               |
| ------------------- | ------------------------------------------ | --------------------- |
| [Files](/files)     | PDFs, manuals, docs                        | Manual re-upload      |
| [Website](/website) | Public docs, help centers                  | Manual re-crawl       |
| [Notion](/notion)   | Internal wikis, frequently updated content | Live (always current) |

## How it Works

1. User asks a question
2. Crow searches your knowledge base for relevant content
3. Agent uses that content to formulate an accurate answer

This is called **RAG** (Retrieval-Augmented Generation).

## Combining Sources

Use multiple sources together:

* **Files** for product manuals
* **Website** for your public documentation
* **Notion** for internal procedures

***

<CardGroup cols={3}>
  <Card title="Files" icon="file" href="/files">
    Upload documents
  </Card>

  <Card title="Website" icon="globe" href="/website">
    Crawl your site
  </Card>

  <Card title="Notion" icon="book" href="/notion">
    Connect workspace
  </Card>
</CardGroup>