# VelvenBot

VelvenBot is Velven's crawler. If it is in your logs, someone listed a page of yours on Velven, or tried to. Velven links to your page and plays it in a frame; it never hosts a copy.

## How to recognise it

Every request Velven makes to your site carries `VelvenBot` in its user agent. The crawler sends:

Crawler:

```text
Mozilla/5.0 (compatible; VelvenBot/1.0; +https://velven.ai/docs/bot)
```

The clip recorder, a Chrome browser, sends:

Clip recorder:

```text
Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/140.0.0.0 Safari/537.36 VelvenBot/1.0 (+https://velven.ai/docs/bot)
```

The addresses they come from change; the `VelvenBot` token does not.

## What it fetches

The crawler fetches only pages someone listed, and only these requests:

- When a page is listed: one GET of the URL, following redirects, reading at most the first 512 KB of HTML for the title, description, preview image and the creator's proof tag. It may also fetch `/.well-known/velven`, an older form of the proof.
- Every 6 hours after that: one GET of each listed page, reading its HTML to check it still answers, still allows framing, and whether it changed enough for a new clip.
- No page is crawled for links.

## The clip recorder

Each tile on Velven shows a 5-second clip of the space. To record it, Velven opens the page in a Chrome browser and plays it for up to a few minutes: after listing, when the page has changed, and when its creator asks for a new clip. Its user agent names VelvenBot, so a block on VelvenBot stops it too.

## Block it

Answer any user agent containing `VelvenBot` with 403 and Velven cannot list the page or record a clip of it. A page that is already listed stays listed, since Velven's check treats a 403 as a page that is up but private; to take it off, see below.

To have a page of yours taken off Velven now, use Feedback on its Velven page, or email [privacy@velven.ai](mailto:privacy@velven.ai). The [privacy page](https://velven.ai/privacy) has the rest.