> For the complete documentation index, see [llms.txt](https://docs.slatecx.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.slatecx.com/guides/add-sources.md).

# Add sources

A source is what allows the SlateCX agent to have meaninful conversations with your prospects.  They are basically the brain behind the SlateCX agent.

Sources live in the Library and are tenant-wide, so any campaign can select any source.

There are different types of sources you can link to SlateCX and more are being added all the time!

The thee most commonly used are:

1. Upload a file. This is very simple. Take a document you have on your computer and manually upload it to SlateCX.
2. Website.  This is probably the most common form of source as SlateCX can simply take your existing webpages and use them as sources.
3. Import HubSpot. If you use HubSpot to store files then you can reference them in SlateCX directly from HubSpot.  You will need to have connected HubSpot first via the Integrations menu.

<figure><img src="https://2844884401-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F5AhClAmEf81WtHS8kM9H%2Fuploads%2Fxp5qMH3JqzLOrihKYbRf%2Fimage.png?alt=media&amp;token=ee710572-5304-411e-86a6-fa1490c39e71" alt=""><figcaption></figcaption></figure>

## Upload a file

1. Open **Library**.
2. Add a source and choose file upload.
3. The client presigns an upload, uploads the file, then finalizes it.

Only usable, finished sources appear in the inventory. Sources still uploading or failed are hidden from the normal list.

## Import a website

Two paths exist:

* **URL import** — queue one or more explicit URLs.
* **Web crawl discovery** — run a discovery run against a site, review discovered URLs, then batch-import the ones you want.

Discovery runs and imports stream progress; the persisted rows remain authoritative.

{% embed url="<https://youtu.be/M5l1I7yhK18>" %}

## Other source types

SlateCX also allows you to use Live external sites as sources.   For example a URL, a YouTube video, a HubSpot form or meeting link, a Walnut demo, or a Storylane demo. Create cards from them explicitly when you want a presentation.

{% embed url="<https://youtu.be/EE9IyWqMl6Y>" %}

## Organize with tags

**Tags** are optional, but they allow you categorize your sources.  You can search in the Libary on tags.

**Tags** are tenant-wide and grouped by trimmed lowercase identity with internal whitespace preserved. Batch tag edits apply one atomic add/remove delta across selected sources.

## Keep sources current

* **Recrawl** — re-fetch crawler-backed HTML page sources.
* **Replace** — stage an immutable replacement document for a document-backed source.
* **Versions** — every source keeps its own version history.

When a source changes, cards backed by it are marked `Source updated`. Campaigns keep using the latest source automatically; cards stay on their saved presentation until you update them.

## Delete a source

You can delete a source but SlateCX will check first that it is not currently in use by a campaign. If it is in use you will need to explicity requested deletion using hte pop-up that is presented.

## See also

* [Sources and cards](/concepts/sources-and-cards.md)
* [Create cards](/guides/create-cards.md)