---
title: Assets
description: Your library of photos, illustrations, videos and PDFs, documented so agents can find and use the right file.
---

> Source: https://www.heyarchetype.com/docs/brand/assets

Assets is your brand's file library: photos, illustrations, SVGs, videos and documents. Each file carries a name, a description, tags and usage notes, so an agent can search for "team photo, outdoors" and get one of yours instead of a stock image.

## Add files

Open **Assets** and press **Add** (choose **Files** if it opens a menu), or drag files anywhere onto the wall. You can add:

- Images, including SVG
- PDFs
- Videos (MP4, MOV, M4V or WebM), up to 500 MB each
- iPhone photos in HEIC or HEIF, which are stored as JPEG so every browser and agent can open them

Archetype reads each file's dimensions and main colors as it lands. If you drop files while a tag filter or folder is open, the new files get that tag or land in that folder.

To make a new picture or clip instead, press **Generate**. See [Images and video](/docs/create/media).

## Let Claude document your files

Right after you upload, Claude looks at each file it can read (images, SVGs and PDFs) and writes a title, a description, tags and usage notes. It reuses the tags your library already has, so filters stay useful instead of splitting into near-duplicates.

To document files that were added earlier, press **Enrich all** above the wall, select files and press **Enrich**, or press **Enrich** on a single file's page. Files you described by hand are left alone. Each description uses AI credits. See [AI credits](/docs/billing/credits).

## Find files

The bar above the wall filters and searches:

- **Type**: All, Image, SVG, PDF, Video or Other.
- **Tags**: pick one or more tags, or show only **Untagged** files. Right-click a tag to rename it or delete it from every file.
- Status, once any file has one.
- **Colour**: pick a color from your library or type a hex, and see files with similar colors.
- **Search files**: matches names, descriptions and tags.

## Folders

To make the first folder, press **Add**, then **Folder**, and name it in place. Folders then sit on a shelf at the top of the wall, which ends in **New folder** for the next one. Drag files onto a folder to file them. A file can be in one folder at a time, and the wall still shows every file, filed or not.

Open a folder to see only its files. Right-click a file inside it to **Use as folder cover** or **Remove from folder**. Deleting a folder keeps its files on the wall.

## Work with a selection

Select files from their checkbox or right-click menu. The bar that appears lets you **Enrich**, **Edit tags**, set a status, **Move to folder**, **Download** or **Delete** them all at once. When the selection includes images, **Download** opens the same format and size options as a single file's **Export** tab.

To tidy the library, open the **…** menu and choose **Scan for duplicates**. Archetype groups identical and near-identical pictures, marks one to keep in each group, and merges the rest only when you press **Merge**.

## A file's page

Click a file to open it. Images and videos show on a stage, and PDFs show page by page. The panel on the right has three tabs.

**Documentation** holds:

- **Details**: format, size, dimensions, colors, and for a PDF its page count and whether its text has been read.
- **Description**, **Tags** and **Usage**: what the file shows, how it's filed, and when to use it.

**Export** saves a copy in another format or size:

- **File format**: PNG, JPEG or WebP, plus SVG for SVG files.
- **Size**: Original, Small (640 px wide), Medium (1,280 px), Large (1,920 px), or a crop for LinkedIn post, Social square, Social portrait, Story or X post.

Archetype never scales a picture up, so sizes larger than the original aren't offered. Exporting several files at once downloads a zip.

**Comments** is a thread on the file, with @mentions.

For a photo or other non-SVG image, the header also offers **Generate from this**: a new picture or clip that starts from it.

## PDFs

Archetype reads the text of every PDF in the background after upload. Agents can then read a product handbook, price sheet or deck page by page and quote it, rather than guessing your facts. A scanned PDF with no text layer is marked as such on its page.

## How agents use your library

Agents start with your tag vocabulary and how many files carry each tag, then narrow by tag or search. Each file comes with its description, usage notes, dimensions, colors and a stable embed URL that is safe to put in generated pages and documents. A PDF also returns its text.

Only files that are **Complete** or have no status are listed for agents. See [Item states](/docs/workspace/states).

Logos are kept apart from the library, in [Foundations](/docs/brand/foundations).

## Next steps

:::cards
- [Images and video](/docs/create/media): Generate new pictures and clips into your library.
- [Templates](/docs/brand/templates): Bind library photos into your layouts.
- [MCP tools](/docs/agents/tools): How agents search and fetch files.
:::
