Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)

A Free Service Turns Any Website Link Into a Three-File Brand Kit for AI Agents

A brand kit is a set of files with a company's colors, fonts, and logo, so that AI doesn't have to invent the visual style from scratch in every chat. A free service builds such a kit from a single website link, producing three files: DESIGN.md, BRAND.md, and SKILL.md. With them, Claude, ChatGPT, and Cursor draw buttons in the brand's green instead of some random blue. Teams without a designer should try the service on their own website.

AI-processed from TG: @notboring_tech (Not Boring Tech); edited by Hamidun News
A Free Service Turns Any Website Link Into a Three-File Brand Kit for AI Agents
Source: TG: @notboring_tech (Not Boring Tech). Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceTG: @notboring_tech (Not Boring Tech) ↗
Editor-in-Chief
Input is simplejust paste the brand's website link
Three filesDESIGN.md, BRAND.md, and SKILL.md for Claude, ChatGPT, and Cursor
Key metric #3It also makes images, banners, and animated posters for social media

If you have asked an AI agent to build a landing page or draw a banner, you know the problem: it does the work neatly, but not in your style. Buttons come out blue instead of your brand green, the font is a random substitute, and the logo looks like someone else's. Every new chat starts with the same explanations. Now a free service has appeared that closes this gap. It takes a link to your website and builds a brand kit: a set of three text files (DESIGN.md, BRAND.md, and SKILL.md) that record the colors, fonts, and logo. You can hand these files to Claude, ChatGPT, or Cursor so they stick to your style without reminders.

What the tool does

The mechanics are simple. You paste your site's address, and the service generates three files. Their names hint at their roles. DESIGN.md is where you would expect design rules: colors, fonts, visual elements. BRAND.md describes the brand itself and its logo. SKILL.md is a file in a "skills" format that agents can load as a ready-made instruction.

According to the description, with these files Claude, ChatGPT, and Cursor can remember the brand's colors, fonts, and logo. The approach will be familiar to developers from context files for coding agents: instead of repeating the rules in every prompt, you put them into the project or settings once, and the agent reads them itself.

The service also creates images, banners, and animated posters for social media. The tool does not stop at describing the brand on paper; it immediately produces finished materials in that style.

Why this matters in practice

Imagine a small startup without a staff designer. It has a website, no brand book, and it needs to publish content every week. Previously, you would have had to write out HEX codes by hand (entries like #1A8F4C, which specify a color exactly), font names, and logo usage rules, then paste them into every prompt. Now that summary can be pulled from a site that is already live: the site itself serves as the source of truth about the brand.

The benefit differs by role:

  • Developer. Places DESIGN.md next to the project, and when generating interfaces Cursor relies on the right colors and fonts instead of inventing its own.
  • Product manager and marketer. Gets banners and posters for social media in a unified style without a separate request to a designer.
  • Founder. Anchors the brand's look in files that can be passed between different AI services, without being tied to any one of them.

The universality of the format matters here. Markdown is readable by both people and any model, so the result is not locked inside a single product. If you switch assistants tomorrow, your brand rules stay with you.

Limitations and what to check

The tool relies on what is on the site. If the site is sloppy, outdated, or uses several different styles, the brand kit will inherit those flaws. Automatically assembled rules are a draft, not a replacement for a full brand book: colors, fonts, and the logo should be checked by eye before you hand them to agents as law.

In addition, the announcement gives no details on how exactly the service extracts styles, or how precisely agents follow the instructions afterward. That has to be evaluated on your own tasks: generate a kit for your site, ask Claude or Cursor to build a couple of screens, and compare the result with the original.

The practical takeaway is simple. The tool is free, takes a couple of minutes to try, and the risk is minimal: at worst you get three text files that you can edit or throw away. If your team regularly generates interfaces and visuals with AI, such a kit can save time on endless corrections of the "now make it in our style" variety. Start with a test project and see how closely the agents hit the brand style.

Story

Key Questions Answered

What the tool does?

The mechanics are simple. You paste your site's address, and the service generates three files. Their names hint at their roles. DESIGN.md is where you would expect design rules: colors, fonts, visual elements. BRAND.md describes the brand itself and its logo. SKILL.md is a file in a "skills" format that agents can load as a ready-made instruction. According to the description, with these files Claude, ChatGPT, and Cursor can remember the brand's colors, fonts, and logo.

Why this matters in practice?

Imagine a small startup without a staff designer. It has a website, no brand book, and it needs to publish content every week. Previously, you would have had to write out HEX codes by hand (entries like #1A8F4C, which specify a color exactly), font names, and logo usage rules, then paste them into every prompt. Now that summary can be pulled from a site that is already live: the site itself serves as the source of truth about the brand. The benefit differs by role: Developer.

What are the key details on Limitations and what to check?

The tool relies on what is on the site. If the site is sloppy, outdated, or uses several different styles, the brand kit will inherit those flaws. Automatically assembled rules are a draft, not a replacement for a full brand book: colors, fonts, and the logo should be checked by eye before you hand them to agents as law. In addition, the announcement gives no details on how exactly the service extracts styles, or how precisely agents follow the instructions afterward.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Hands-On AI Engineering · Hamidun Academy

Build & Ship Production AI Agents Without Boilerplate Code

Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.

Start 72h Free Trial →Vibe-Coding Intensive
What do you think?
Loading comments…