ives

In memory of Jony Ive — who showed us that design is how it works.

A complete design philosophy and practical UI knowledge system for building exceptional interfaces. 13 composable skills that give AI coding agents the ability to reason about design — not just follow rules.

Works with: Claude Code, Cursor, Codex, OpenCode, Amp, and any tool supporting the Agent Skills open standard.

What makes this different

Every other design skill in the market falls into one of four categories:

• Rule-based linters — "your code violates these design rules"
• Post-processing fixers — "clean up the slop after generation"
• Design memory systems — "remember and enforce your choices"
• Reference databases — "query this catalog of good design"

ives is the layer underneath all of them: design reasoning. It teaches the agent how to think about design contextually — not just what to do, but why, and when it matters.

A real designer doesn't apply the same rules to a dense data dashboard and a luxury landing page. They reason about information density, user intent, visual hierarchy relative to content. This skill set encodes that reasoning.

Architecture

Tier 1: Craft Process (how to work)

Skill	Purpose
craft-philosophy	Foundational mindset: noticing, uncommon care, less but better, live tuning
design-critique	Structured 4-lens review: visual design, interface design, interaction consistency, user context
interface-architecture	Structural decisions: platform defaults, information architecture, visual language
conceptual-exploration	Creative process: range (breadth of ideas) and depth (pushing a direction)
quality-facets	Quality measurement: define, evaluate, and track what "good" means for your product
refinement-playbook	Iteration process: remove redundancy → simplify → tighten → consistency → hierarchy → craft

Tier 2: Design Knowledge (what good looks like)

Skill	Purpose
typography	Type sizes, weight as meaning, hierarchy, system fonts, the rule of four
layout-spacing	Box model, grids, negative space, alignment, density, spacing systems
color	Palettes, contrast/a11y, gray systems, dark UI, structural vs interactive color
style-direction	Corner radius, shadows, borders, opacity, buttons, design direction
imagery	Photos, icons, illustrations, app icons, dynamic images
elements-patterns	Navigation, inputs, forms, lists/cards, modals, tables, detail views
tactics	No-stress process, wireframes, responsive, platform guidelines, handoff

Installation

Claude Code (plugin)

/plugin marketplace add your-username/ives

Claude Code (manual)

git clone https://github.com/your-username/ives.git
cp -r ives/skills/* ~/.claude/skills/

Cursor / Other tools

Copy the skill files into your project's prompt or context directory. The SKILL.md format is compatible with 40+ AI coding tools that support markdown-based instructions.

How it works

The skills use progressive disclosure — only the skill name and description are loaded at startup (~100 words each). The full skill body is loaded only when relevant. Reference files with deep knowledge are loaded on-demand.

This means all 13 skills add minimal overhead to your context window while being available whenever design decisions arise.

Typical flows

Building something new:

1. craft-philosophy loads as baseline mindset
2. conceptual-exploration helps explore before committing
3. interface-architecture guides structural decisions and visual language
4. typography, color, layout-spacing provide domain knowledge
5. design-critique runs self-critique before presenting

Improving something existing:

1. design-critique identifies specific issues
2. quality-facets measures and prioritizes
3. refinement-playbook systematically resolves
4. Domain skills (typography, color, etc.) provide specific guidance

Quick review:

1. design-critique runs the 4-lens framework
2. quality-facets if structured evaluation is needed

Composability

ives is designed to work alongside other tools:

• Use with rams.ai for automated accessibility rule checking
• Use with ui-skills for post-generation fixes
• Use with interface-design for design memory/persistence
• Use with Anthropic's frontend-design skill for aesthetic steering

This skill set provides the reasoning layer that makes all of those tools more effective.

Philosophy

"Sometimes magic is just someone spending more time on something than anyone else might reasonably expect." — Penn & Teller

Most AI-generated interfaces stop at level 3 on a 1-10 quality scale. Not because the AI isn't capable of level 7+, but because nobody taught it to push further. ives encodes the mindset, knowledge, and process to close that gap.

Contributing

This is an open-source project. Contributions welcome:

• Additional domain knowledge (animation, micro-interactions, data visualization)
• Platform-specific guidance (SwiftUI, React Native, Flutter)
• Real-world examples and case studies
• Translations

License

MIT