AI Agent Hub
Back to skills
🎨

Frontend Design

Design & Media Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please install @org-02qudk26/frontend-design1 according to https://skillhub.cn/install/skillhub.md.

About this skill

Problem to Solve

When models generate frontend pages, the usual issue is not whether a page can be rendered, but whether it looks like one of the same defaults: a warm cream background with a high-contrast serif and terracotta accent, a near-black layout with acid-green accents, or a broadsheet grid with hairline rules. If those choices are not tied to the subject, the result feels templated. This skill asks the model to write down aesthetic decisions before producing a page that merely looks like a demo. It targets front-end work where HTML, CSS, and copy need to be produced together, with the subject, audience, and single page job fixed early.

How It Works and Where It Fits

  • Ground the subject and tokens: express the palette as 4–6 named hex values, assign distinct roles to display, body, and utility typefaces, and compare directions with one-sentence layouts and ASCII wireframes.
  • Critique before building: check whether the plan is only a reusable default; if so, revise it and state the reason, then implement the confirmed plan.
  • Handle implementation details: avoid selector conflicts such as .section overriding .cta, and respect mobile layout, keyboard focus, prefers-reduced-motion, and action-oriented copy like Save changes.
  • Boundaries: it is best for design judgment and expressive front-end pages, not for low-variance admin panels, standard component-library assembly, or fully locked brand systems; when the brief pins the visual direction, the brief wins.

Use Cases

  • Drafting a homepage concept by fixing subject, palette, type roles, and one page job before producing a reviewable layout direction.
  • Turning a rough brief into a front-end design direction with color tokens, display and body typefaces, ASCII wireframes, and a signature element.
  • Reviewing an AI-generated landing page for templated defaults such as cream serif or acid-green-on-black, then justifying changes.
  • Writing front-end copy that replaces system terms with user-visible actions like Save changes and keeps toast states consistent.

Best For

  • Front-end engineers responsible for homepage and campaign pages who need to turn requirements into subject-grounded visual and copy directions.
  • Product designers prompting large models for front-end design who need self-critique of defaults before token and wireframe output.
  • Copywriters maintaining landing-page text who need consistent, actionable labels, errors, empty states, and toast messages.
  • Tech leads reviewing AI-generated pages for templated style, structure, motion, and CSS selector risk.