Making covers and mockups with your AI

Updated Nov 6, 2026

With the AI Skill Pack, your own AI makes your cover. In ChatGPT, it paints the artwork with no words in it, and the connector sets your title in the Library's own type over it. If your AI can't make images, as in Claude, your files get the Library's type cover in your color. Entrepedia doesn't run an image model for you.

How a Library cover is made

The Master Library's covers are made in two parts, and the skill copies that:

The artwork. One everyday object a stranger can name in a second, with one detail that carries the promise. For example, an invoice with a tick stamp sliding into a wallet for a book on getting paid on time.
The type. The title, subtitle and your name are set by the Library's template, not painted by the image model. That keeps the spelling exact and the covers consistent.

Before painting anything, the skill writes three cover ideas and checks each one: can a stranger name the object in a second, does it carry the promise, does it need explaining, is it too cute? Ideas that fail are replaced.

In ChatGPT (it can make images)

Ask for a cover, for example: "Make a cover for my planner in my colors."
ChatGPT picks one of the Library's four looks (dark, light, color or classic) to match your brand.
It paints the artwork only, with no text anywhere in it, and checks the picture for stray letters.
It passes the artwork to the connector, which sets your title over it and returns the PDF, the Word file and the finished cover as a picture.
It reads every word on the finished cover against what you gave it.

In Claude (it usually can't make images)

Claude says so at the start. Your files then get the Library's type cover: a strong field of your color with your title in the Library's faces. If your brand needs its own typeface or logo on the cover, Claude can code a cover as an SVG drawing instead and pass it to the connector.

Mockups for your listing

From the finished cover, the skill makes four to six listing images: the product on a tablet, on a phone, as a printed book or binder on a desk, a spread of inside pages, and a "what's inside" image. For bundles, everything together.

With an image model: your AI makes them, using the finished cover as the reference.
Without one: your AI writes a short brief for each scene that you can run in any image tool.

The skill asks where you'll sell, because Amazon, Etsy, Gumroad and Instagram each want different image sizes.

Rules the skill follows

No lettering painted by the image model.
Your title, subtitle and name spelled exactly as you gave them.
Mockups show the real product, with no fake reviews, badges or "bestseller" seals.
No other brands, characters, real people's faces or someone else's artwork.
A reminder that Amazon KDP asks whether a cover is AI-generated and Etsy asks sellers to disclose AI use. The answer is yours to give.

Other cover tools in your library

The Master Library also includes its own Cover Generator and Mockup Studio, under Create in the Library menu. They work on their own and don't need the AI Skill Pack.