Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Design Setup

ASecurity

Complete end-to-end web design setup & prototyping framework faithful to Chase AI's methodology with production robustness gates. Automatically enforces design-rules (micro-UI typography hacks, 4pt/8pt grid, dark mode surface elevation, 4-state buttons) and legend mastery principles (unconventional inspiration, signature calling cards). Interactively guides users to cultivate design taste, enforces preflight checks, decision gates, exploratory code isolation, 4-part prompts, 3-phase build seq...

2 stars
0 votes
0 copies
0 views
Added 10/5/2026
ai-agentsgonextjsrailsgit

Works with

claude codeterminalclimcp

Security Analysis

A100/100

Scanned 10/5/2026

$npx -y skills add xenodeve/xeno-skills --skill design-setup --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Design Setup?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Design Setup
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/xenodeve-design-setup/badge)](https://www.skillsdirectory.com/skills/xenodeve-design-setup)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: design-setup
description: Complete end-to-end web design setup & prototyping framework faithful to Chase AI's methodology with production robustness gates. Automatically enforces design-rules (micro-UI typography hacks, 4pt/8pt grid, dark mode surface elevation, 4-state buttons) and legend mastery principles (unconventional inspiration, signature calling cards). Interactively guides users to cultivate design taste, enforces preflight checks, decision gates, exploratory code isolation, 4-part prompts, 3-phase build sequence, and token commit finalization.
triggers:
  - /design-setup
  - design-setup
  - design setup
  - initialize web design
  - setup website UI
  - overhaul web design
---

# Complete Web Design Setup Framework (`design-setup`)

This skill guides AI agents and users through setting up bespoke, high-taste web UIs and landing pages from scratch while systematically eliminating **AI Slop** (generic dark slate modes, cyan/purple gradients, default Inter fonts, 3D blobs, uniform 3-column cards, and cliché copy).

*Automatic Skill Integration:* During UI component and layout generation, this skill **automatically enforces micro-design rules from `design-rules`** (e.g., `-2%` headline tracking hack, 4pt/8pt grid spacing, dark mode surface elevation, 4-state buttons, semantic colors, and progressive backdrop blur overlays).

---

## ⛔ 0. Environment Preflight Check (Mandatory Pre-Run Audit)

Before initiating the design setup workflow, the agent MUST run a brief preflight check to verify that all necessary dependencies are in place:

1. **Skills Audit:** Verify if **Impeccable Skill** (`impeccable.style`), **Taste Skill** (v2), **UI/UX Pro Max**, and **`design-rules`** are active.
2. **MCP Connectivity:** Verify that **Higsfield MCP** CLI (`higsfield.ai`) is active and authenticated if background hero imagery/video is required.
3. **Dev Server Status:** Ensure local dev server (e.g., Vite, Next.js) and preview environment are ready.
4. **Inspiration Check:** Confirm if the user has provided reference screenshots, URLs, or an Inspiration Library brief.

*Action:* If any dependency is missing, prompt the user immediately before starting execution to prevent getting stuck midway.

---

## 🚫 1. Anti-Slop Guardrails & Micro-Rules Enforced

Enforce these strict negative constraints on EVERY UI generation task:

* ❌ **NO Generic Color Palettes:** Never default to `#0f172a` slate dark mode with cyan (`#06b6d4`) to purple (`#a855f7`) gradients.
* ❌ **NO Font Defaulting:** Never use `Inter` font without an explicit font scale and intentional pairing (e.g., Serif headline + Monospace label).
* ⚡ **Headline Pro Typography Hack (via `design-rules`):** Apply `-2%` to `-3%` letter-spacing (`tracking-tight`) and `110%-120%` line-height (`leading-tight`) on large heading text.
* 📐 **4pt/8pt Spacing Grid (via `design-rules`):** All padding, margin, and gap values MUST be exact multiples of 4 or 8.
* 🌙 **Dark Mode Surface Elevation (via `design-rules`):** Create depth in dark mode using lighter background surface colors (`#09090b` ➔ `#18181b` ➔ `#27272a`), NOT black drop shadows.
* 🔘 **4-State Buttons & Icon Line-Height (via `design-rules`):** Render Default, Hover, Active, Disabled button states. Match icon size to text line-height.
* ❌ **NO Cliché Elements:** No 3D glowing glassmorphism blobs, uniform 3-column feature cards with gray `border-slate-800`, or floating neon rings.
* ❌ **NO Vague Copywriting:** No generic AI headings like *"Ship products that think for themselves"*. Use authentic, product-specific copy.

---

## 🧭 2. Step 1: Cultivate & Curate Taste (Inspiration & Reference Protocol)

AI has no inherent taste and defaults to "regression to the mean." Before writing code, the user MUST cultivate taste and gather high-level design inspiration.

### 2.1 Standard & Unconventional Inspiration Sources
When asking the user for inspiration or references, provide these primary and unconventional sources:

#### Standard Sources:
1. **Dribbble (`dribbble.com`):** Search `web design` + filter by `Popular` to discover high-level landing page layouts.
2. **Pinterest (`pinterest.com`):** Search `web design` or `UI hero section` to find unconventional hero compositions.
3. **Twitter / X:** Follow active UI/UX design creators for cutting-edge, experimental UI patterns.

#### 🌟 Unconventional Sources (Look Where No One Else is Looking):
4. **Typographic Posters (`typographicposters.com`):** High-impact poster layouts & unconventional type arrangements.
5. **Savee (`savee.it`):** High-curation design community free of ad noise.
6. **Cross-Disciplinary Inspiration:** Architecture movements, design history books, museum exhibition layouts (synthesizing non-web fields creates ultra-original UIs).

### 2.2 Inspiration Library Web App & Copy Brief / Copy Image Prompt
To organize inspiration, the user can have Claude Code build a simple **Inspiration Library Web App** that groups screenshots into aesthetic categories and provides:
* **Design Vocabulary & Keywords:** Tags describing the visual style (e.g., *Voxel rendered landscape, Monolithic scale, High-contrast index*).
* **Copy Image Prompt Button:** Copies the image prompt to generate background hero imagery.
* **Copy Brief Button:** Copies the foundational design brief to paste directly into Claude Code.

*Agent Action:* Prompt the user:
> *"Before we build, do you have reference screenshots, URLs, or a copied Brief from your Inspiration Library? If not, tell me which aesthetic family (e.g., Print Tech, Dither Mono, Vast Quiet, Classical Remix) fits your vision!"*

---

## 🧰 3. Step 2: Equip External Tools & Calling Card Superpowers

### 3.1 ⚠️ Beware Going Down the Tool Rabbit Hole
Do NOT fall into the trap of installing hyper-prescriptive skills that promise to solve all design problems with one click. Narrow, prescriptive skills only produce one rigid output. Favor **flexible, non-prescriptive tools** (Impeccable, Taste Skill, Higsfield MCP) while utilizing popular UI repositories like UI/UX Pro Max for extra baseline rules.

### 3.2 External Ecosystem & Calling Card Matrix

| Tool / Skill | Access / Command | Primary Function |
| :--- | :--- | :--- |
| **`design-rules`** | Automatic Integration | Micro-UI engineering rules (Headline hack, 4pt/8pt grid, dark mode elevation, button states). |
| **Impeccable Skill** | `impeccable.style` | Audits 46 anti-slop patterns across 7 dimensions (*Typography, Color, Spatial, Responsiveness, Interaction, Motion, UX Writing*). Provides 23 commands (e.g., `/boulder`, `/clarify`, `/overdrive`) and Live Mode on Dev Server. |
| **Taste Skill (v2)** | `npx skills add Leonxlnx/taste-skill` | Enforces custom typography scales, spacing systems, and motion curves (`tasteskill.dev`). |
| **UI/UX Pro Max** | `github.com/nextlevelbuilder/ui-ux-pro-max-skill` | Popular comprehensive UI/UX design skill repository providing structured design guidelines, component patterns, and prompt rules. |
| **Higsfield MCP** | `higsfield.ai` (MCP CLI) | Model Context Protocol integration for generating custom background hero images (e.g., GPT Images 2) and ambient video loops (e.g., Seed Dance). |
| **21st.dev** | `21st.dev` | Component-level inspiration (Buttons, Cards, Pricing tables). Allows copying precise component prompts. |
| **Calling Card Tech** | GSAP / Spline 3D / WebGL | Signature interactive superpower technique (e.g., GSAP ScrollTrigger, 3D interactive assets) creating a distinct visual calling card. |

---

## 📁 4. Exploratory Code Scoping Policy

To prevent cluttering production codebase with abandoned components during design exploration:
* **Exploration Directory:** Place all multi-style explorations in isolated exploration routes/folders (e.g., `/design-exploration/style-1`, `/design-exploration/style-2`).
* **Production Route Protection:** Do NOT overwrite main production routes (e.g., `app/page.tsx` or `index.html`) until the user explicitly approves the final design direction.

---

## 📐 5. The 4-Part Prompt Structure

Every design setup prompt passed to the agent SHOULD contain these 4 components:

```markdown
1. Aesthetic Family: [e.g., Print Tech Paper, Dither Mono, Vast Quiet / Cinematic, Classical Remix, or Custom Hybrid]
2. Reference Image / URL: [Screenshots, Live URL, or Copied Brief from Inspiration Library]
3. Intent & Target Audience: [Product purpose, target audience, primary CTA e.g., Book Demo]
4. Guardrails: [Explicit constraints e.g., No purple gradients, no 3D blobs, no default Inter font]
```

---

## 🔄 6. Step 3: The 3-Phase Build Sequence & Hard Decision Gates

Never attempt a "One-Shot" design setup. Execute this exact iterative build sequence with **Hard Decision Gates**:

### Phase 1: Cast a Wide Net (5 Styles Comparison Grid)
Prompt the agent to generate **5 distinct versions** of the page across 5 different aesthetic styles in isolated exploration routes:
* Example Prompt: *"Setup a landing page for [Product Name], [Intent & CTA]. Guardrails: no purple gradients, no 3D blobs, no Inter font. Create 5 versions of the page in 5 different directions in /design-exploration/ (e.g., Print Tech Paper, Vast Quiet, Dither Mono, Classical Remix, Industrial Minimal)."*
* *Micro-Rule Enforcement:* Apply `design-rules` headline tracking (`-0.025em`) and 4pt/8pt grid spacing across all 5 versions.

> 🛑 **HARD GATE 1:** STOP. Do NOT proceed to Phase 2 until the user explicitly selects ONE direction from Phase 1.

---

### Phase 2: Explore Body Layout Variants (1 Style ➔ 3 Variants)
Once the user selects a preferred style from Phase 1 (e.g., *Vast Quiet*):
* Prompt the agent: *"Let's go with the Vast Quiet version. Generate 3 versions of that aesthetic for me, namely changing the body layout formats (e.g., Minimal Vertical Stack, Quiet Ledger with sticky left index, Framed Border Structure)."*
* *Micro-Rule Enforcement:* Render components using 4-state buttons, ghost buttons, and dark mode surface elevation.

> 🛑 **HARD GATE 2:** STOP. Do NOT proceed to Phase 3 until the user explicitly selects ONE body layout variant from Phase 2.

---

### Phase 3: 2-Pass Hero Image Generation via Higsfield MCP
Execute a 2-Pass image generation process using Higsfield MCP for the background hero asset:
1. **Pass 1 (Composition Options):** Prompt the agent: *"Let's nail that hero image. Give me 4 different high-quality 2K image compositions that fit our quiet aesthetic hero section using Higsfield MCP."* (e.g., *Aerial view, Crag mountain, Watercolor, Cloud Sea*).
2. **User Selection 1:** Select 1 composition (e.g., *Option 4: Cloud Sea*).
3. **Pass 2 (Color & Lighting Options):** Prompt the agent: *"Let's go with Option 4. Create multiple versions of that adding a splash of color (e.g., Dawn, Golden Hour, Alpenglow, Duotone)."*
4. **User Selection 2:** Select the final hero image (e.g., *Alpenglow*).

> 🛑 **HARD GATE 3:** STOP. Do NOT implement the final production page until the final hero asset is selected by the user.

---

## 🛠️ 7. Tweaks Bar & Page Motion Directives

### 7.1 Natural Prompting for Dev Server Tweaks Bar
Instead of guessing fixes in terminal text, prompt the agent to add an interactive **Tweaks Bar** to the preview page:

> **Prompt:** *"Can we sort of mimic what happens inside of Claude Design and add a tweaks bar that pops up on this dev server so I can change a number of things? Whether that's font size, font type, accent colors, hero asset variations, motion speed, reveal distance, or element weight. Go pretty aggressive with what you offer me."*

The agent will inject a floating control overlay on the dev server allowing visual tweaking of:
* Heading & Body Fonts (with Italics toggle and Font Size sliders)
* Hero Image Asset Variations toggle
* Motion Speed, Reveal Distance, and Element Weight / Stagger Delay
* Accent Colors and Spacing

### 7.2 Page Loading Weight & Transition Directives
* **Hero Transition:** The hero background MUST smoothly fade out into the body section upon scrolling (no sudden color cuts).
* **Page Loading Weight:** Page loading **MUST feel heavy and carry weight**. Use staggered reveal animations so elements load sequentially into place with tangible weight.

---

## 🏁 8. Finalization & Production Commit Phase

Once the design is approved by the user, execute this finalization checklist:

1. **Commit Selected Tweaks Values:** Extract the user's final selected values from the Tweaks Bar (fonts, accent colors, spacing, stagger delay) and commit them permanently into production design tokens (e.g., `tailwind.config.js` or global CSS variables).
2. **Clean Up Exploratory Code:** Remove the temporary Tweaks Bar overlay and delete unused exploration routes (`/design-exploration/`).
3. **Promote to Production Route:** Move the finalized page into the main production route (e.g., `app/page.tsx` or `index.html`).
4. **Quality Checks:**
   * Verify micro-design compliance against `design-rules` (headline tracking, button states, dark mode elevation).
   * Verify responsive layout across Mobile, Tablet, and Desktop.
   * Verify basic accessibility (contrast ratios, heading hierarchy).
   * Respect `prefers-reduced-motion` for animations.
   * Compare final screenshot against the original selected inspiration direction.

---

## 📚 9. Source Video References & Transcripts

For complete untruncated transcripts and raw video research:
* 📄 [01_chase_ai_anti_slop_web_design.md](../references/01_chase_ai_anti_slop_web_design.md) — Chase AI
* 📄 [05_chris_mccoy_5_web_design_skills.md](../references/05_chris_mccoy_5_web_design_skills.md) — Self-Made Web Designer
* 📄 [06_chris_mccoy_6_habits_of_design_legends.md](../references/06_chris_mccoy_6_habits_of_design_legends.md) — Self-Made Web Designer
* 📄 Full Reference Index: [README.md](../references/README.md)

Attribution

xenodevexenodeve
View sourceSee grades on GitHubMore from xenodeve →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698621 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →