Frontend design: distinctive, intentional visual design guidance (official Anthropic plugin)
Design-lead style guidance to escape templated AI-generated looks — typography, restraint, two-pass plan-then-build process
Curated by Skill Harbor — the official @anthropics plugin that makes an agent design like a design-studio lead: deliberate, opinionated choices about palette, typography and layout that read as a specific point of view, never a template. It grounds every design in the brief's subject matter, runs a two-pass process (plan a compact token system — color, type, layout, principles — in ASCII wireframes, then critique it against the brief before writing a single line of code), calls out the five visual clichés of AI-generated pages (warm-cream with terracotta, acid-green-on-black, broadsheet hairlines, the SaaS-card kit, ALL-CAPS eyebrow chrome), and demands restraint — one memorable element, everything else quiet — plus responsive, accessible, keyboard-navigable output and copy written for end users, not systems. Apache-2.0 licensed. Honest notes: this is taste guidance, not components — it won't write your app for you; and its "avoid these defaults" advice should bend to the brief, which always wins. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Quick answer
This is Anthropic's official frontend-design plugin: guidance that makes an agent design like a design-studio lead. Deliberate choices about palette, typography and layout, a two-pass plan-then-build process, and restraint — one memorable element, everything else quiet — instead of the templated look of AI-generated pages.
At a glance
| What it is | Taste guidance, not components: a set of design-lead principles and a working process your agent applies when it designs and builds interfaces for you. |
|---|---|
| Who it's for | People who build pages or apps with an AI agent and keep getting the same generic look — and want output with a specific point of view instead. |
| What you give it | A brief: the page or interface to design, and its subject matter. The process grounds every design decision in that brief — the brief always wins. |
| What you get | A compact plan first — a token system for color, type and layout, sketched in ASCII wireframes and critiqued against the brief — then the build: responsive, accessible, keyboard-navigable output with copy written for end users. |
| Works with | Muse, installed from this page with the copy-paste install prompt. It is an official Anthropic plugin, Apache-2.0 licensed. |
| Cost | Free |
| Time | A few minutes to install. The two-pass process adds a short planning step before the build, instead of designing while coding. |
Example — a landing page that doesn't look AI-generated
An example of the process the plugin documents, applied to a page brief:
- You give the brief: a landing page for a small bakery, warm and plain-spoken. The agent first plans a token system — palette, type, layout — and sketches it in ASCII wireframes.
- Before any code, it critiques the plan against the brief and against the clichés it is taught to avoid — no warm-cream-and-terracotta default, no generic SaaS card kit — and picks the one element meant to be memorable.
- Only then does it build: the chosen direction carried through, everything else kept quiet, and the result checked for responsive behavior and keyboard navigation.
Install
Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Nov 18, 2026📌 Pinned version: commit b860d6fStep 0 — before installing anything: read this install prompt, summarize in a few lines what this build does and what installing it involves, then ask me to confirm before you proceed. Prerequisites: nothing to install — the skill is design guidance your agent follows when building or reshaping UI Install "Frontend design: distinctive, intentional visual design guidance (official Anthropic plugin)" for me. It gives my agent @anthropics's official design-lead approach: ground every design in the brief's subject matter, work in two passes (plan a compact token system in ASCII wireframes, critique it against the brief before writing code), avoid the five clichés of AI-generated pages, spend boldness in one place, and ship responsive, accessible, keyboard-navigable UI with end-user copy. Apache-2.0 licensed. Repository: https://raw.githubusercontent.com/anthropics/claude-plugins-official/b860d6f2554f145fe1e7ac4cdcabca0120d14030/plugins/frontend-design/skills/frontend-design/SKILL.md Pinned version: install the code at commit b860d6f2554f145fe1e7ac4cdcabca0120d14030 exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Fetch the SKILL.md file (and any helper files) from the repository path into a temporary folder 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code, credentials only via the secure vault, allowed hosts declared in the SKILL.md. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "frontend-design". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. nothing — just ask the agent to apply it on the next UI task). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Saved to your recent installs. Find it anytime on /connect.
🔒 Security, provenance and what this touches
The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?
Verification certificate
Dated facts about what was checked for this listing, and how. This is not a safety certification.
- Pinned version
- b860d6f · 2026-10-13
- Sandbox install test
- Install ran in our test sandbox · Nov 18, 2026 · Observed during the install test: registry.npmjs.org
- Creator identity
- Not verified
- Fingerprint watch
- Last checked · Nov 19, 2026 : no change detected in the files linked by this install prompt
- Install prompt analysis
- 2026-10-13 : No commands, network calls, packages, file writes or API keys detected.
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Questions
How do I install a skill?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.