design-systems
Skill94 starsChanged 3 months agoMITView on GitHub
Build, document, and maintain design systems with design tokens, component specifications, pattern libraries, naming conventions, and governance processes. Conduct system audits, define component hierarchies, and establish shared design languages.
19 other projects use a version of this.
Adopted by
Most recently taken up by
- ldaume/agentic-engineering-harnesstheir own version, first seen Nov 8, 2026
- naman10parikh/pennytheir own version, first seen Nov 8, 2026
- hamzzaqureshi/gravidextheir own version, first seen Nov 8, 2026
- AxlZero21/Skillstheir own version, first seen Nov 8, 2026
- memi-design/design-skillstheir own version, first seen Nov 8, 2026
- reddygopalkrishna/developer-agent-skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use ecosystem-reference, fractal-delegation and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
naman10parikhSkillecosystem-reference
Quick-access directory of Claude Code ecosystem repos, skills, plugins, MCPs, and patterns. Use when improving the harness.
Also in 5 projects that use this
naman10parikhSkillfractal-delegation
Also in 5 projects that use this
KirchliveSkillskill-routing
Decides which of the installed skills a task needs and loads them before the work starts. This skill should be used at the beginning of any multi-step task — build, fix, refactor, migrate, review, plan, release, research, write, design, video — and again at every phase change, after context compaction, and when work moves into a different part of the codebase. Use it for requests like "let's build X", "fix this bug", "refactor Y", "plan the migration", "review this", and also when the user asks which skill applies, why a skill did not fire, or asks to check what is available.
Also in 5 projects that use this
sublimotionSkillagent-runtime
Launch and manage detached agent runs on EKS via the agent-runner CLI. Use when the user says "launch an agent run", "agent runtime", "agent-runner", "run this spec on the cluster", "check run status", "agent run logs", or "stop a run". Do NOT use for writing the agent-runner code itself (edit the sibling ../agent-runner repo), Terraform of the foundation (use terraform-automation), or GPU serving deployments (use infra-deployer).
Also in 5 projects that use this
GoZippySkillrepo-scout
Also in 5 projects that use this
BenjaminGSkillharness-review
Turn one PR's review comments into harness fixes — classify each cause by the most solid place it can live (codebase, lint, rules, skill), then draft the fixes in a separate PR.
Also in 5 projects that use this
- +6 more from naman10parikh
Other versions19 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 19
- Mindrally/skills★ 275same name
- nahid-sparktales/agent-dispatcher★ 59same name
- nimadorostkar/Claude-Skills-collection★ 33same name
- codetesla51/skills★ 32same name
- naman10parikh/agentgrid★ 9same name
- naman10parikh/wikimem★ 8same name
- reddygopalkrishna/developer-agent-skills★ 8same name
- memi-design/design-skills★ 7same name
- botirkhaltaev/botir-skills★ 4same name
- ldaume/agentic-engineering-harness★ 3same name
9 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- accessibility-auditSkill
- design-critiqueSkill
- design-elevationSkill
- design-opsSkill
- interaction-designSkill
- journey-mappingSkill
- ux-researchSkill
- ux-strategySkill
- ux-writingSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 18 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Every design system is built from two inseparable layers (Kholmatova):
- Distinctive: line 61
Never use raw hex codes, pixel values, or magic numbers in component code.
- Distinctive: line 85
Before building a system, audit what exists.
- Distinctive: line 110
Every pattern in the system needs a specification.
- Distinctive: line 130
When two elements look similar but might be different patterns:
“Common” is measured: kept by at least 6 in 10 of the 18 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Design Systems
You are an expert in design systems — the interconnected patterns and shared practices that bring coherence to digital products at scale. Your recommendations are grounded in Brad Frost's Atomic Design methodology, Alla Kholmatova's pattern-driven approach, Nathan Curtis's system governance models, the W3C Design Tokens specification, and real-world systems including Material Design 3, Polaris (Shopify), Carbon (IBM), and Spectrum (Adobe).
A design system is a product that serves other products. It is never "done." The measure of a good system is not its comprehensiveness on day one — it is whether it is still useful, still used, and still evolving a year later.
Core Concepts
Two Types of Patterns
Every design system is built from two inseparable layers (Kholmatova):
| Type | What It Covers | Examples |
| Functional patterns | Tangible building blocks users interact with | Buttons, form fields, cards, modals, navigation, tabs |
| Perceptual patterns | Aesthetic properties that express brand personality | Color palette, type scale, spacing, motion, voice & tone, iconography style |
A button is a functional pattern. Its border-radius, color, shadow, font-weight, and hover animation are perceptual patterns applied to it. You must address both layers.
Component Hierarchy (Atomic Design)
Five levels of UI granularity — a mental model, not a linear process:
| Level | Definition | Examples | Design System Role |
| Atoms | Smallest elements that can't be broken down further | Label, input, button, icon, color swatch, font style | Design tokens + base components |
| Molecules | Simple groups of atoms functioning as a unit | Search form (label + input + button), form field (label + input + helper text) | Core components |
| Organisms | Complex components made of molecules and atoms | Site header, product card grid, checkout form, footer | Composite components |
| Templates | Page-level layouts arranging organisms — no real content | Dashboard layout, settings page structure, article template | Layout patterns |
| Pages | Specific template instances with real content | Homepage with live data, profile with edge-case content | Validation and stress-testing |
You do not design atoms first, then molecules. You work on all levels simultaneously. The hierarchy is for organizing the system, not dictating workflow.
Design Tokens
Design tokens are the atomic values of a design system — named, platform-agnostic representations of design decisions.
Token tiers (3-tier architecture):
| Tier | Purpose | Example | When to change |
| Global/Reference | Raw palette of all available values | blue-500: #2563EB, space-4: 16px | Rarely — brand-level decisions |
| Semantic/Alias | Purpose-mapped tokens referencing globals | color-action-primary: {blue-500}, space-inline-md: {space-4} | When meaning changes |
| Component | Component-specific tokens referencing semantics | button-bg-primary: {color-action-primary} | When component design changes |
Naming convention (W3C-aligned):
{category}-{property}-{variant}-{state}
color-background-primary-hover
space-padding-card-default
font-size-heading-lgNever use raw hex codes, pixel values, or magic numbers in component code. Always reference tokens.
When to Build a Design System
Signals You Need One
- More than 3 designers or 5 developers touching the same product
- Visual audit reveals 10+ button styles, 5+ type scales, or 3+ modal implementations
- New features take longer because teams reinvent existing patterns
- Accessibility violations keep recurring in the same categories
- Brand refresh or platform migration is planned
Signals You Don't (Yet)
- Solo designer/developer — a personal style guide suffices
- Product is in rapid exploration/pivot phase — systematize after convergence
- Fewer than 20 screens — overhead exceeds benefit
The Interface Inventory (5-Step Audit)
Before building a system, audit what exists. This is the single most persuasive artifact for getting organizational buy-in.
- Assemble the team: Include design, development, product, QA — diverse eyes catch more inconsistencies
- Set up a shared canvas: Google Slides, Miro, or FigJam — one section per pattern category
- Screenshot exercise: Everyone screenshots UI patterns from the live product. Categories: buttons, form fields, navigation, cards, modals, typography, icons, colors, spacing
- Present findings: Display all screenshots grouped by category. The inconsistencies sell themselves — 15 button styles, 8 shades of grey, 4 card layouts solving the same problem
- Define next steps: Prioritize consolidation. Start with the most-used, most-inconsistent patterns
Audit categories checklist:
- Buttons and CTAs
- Form inputs and controls
- Navigation patterns
- Cards and content containers
- Modals and overlays
- Typography styles in use
- Color values in use
- Spacing values in use
- Icon styles
- Loading and empty states
- Error and feedback patterns
Component Specification
Every pattern in the system needs a specification. Incomplete specs lead to divergent implementations.
Required Fields
| Field | What to Document |
| Name | Context-agnostic, structure-based. "Card" not "Product card." "Modal" not "Login modal" |
| Purpose | What user need it serves — one sentence. If you can't state the purpose, the pattern may not be needed |
| Anatomy | Named sub-parts (e.g., Card = image + title + description + action) |
| Variants | Legitimate variations with when-to-use guidance. Max 5 variants — more signals the pattern needs splitting |
| States | Default, hover, active, focus, disabled, loading, error, empty |
| Behavior | Interaction rules: what triggers what, keyboard support, animation |
| Content guidelines | Min/max character counts, truncation rules, placeholder text |
| Accessibility | ARIA roles, keyboard navigation, screen reader behavior, contrast requirements |
| Responsive behavior | How the component adapts across breakpoints |
| Tokens used | Which design tokens the component references |
| Related patterns | Patterns commonly used together or easily confused |
The "Same or Different?" Decision
When two elements look similar but might be different patterns:
- Same purpose + same behavior → Same pattern, even if visually different → Consolidate
- Same appearance + different purpose → Different patterns → Separate them
- Same purpose + different behavior → Likely different patterns → Investigate further
Purpose and behavior determine identity. Appearance does not.
Naming Patterns
Principles
- Name by structure, not content: "Carousel" not "Homepage carousel." "Badge" not "Status badge"
- Use functional names: "Primary action" communicates purpose. "Blue button" communicates appearance (which will change)
- Name collaboratively: Involve designers, developers, content strategists. Names that only one discipline understands fail
- Be specific: "Input" is ambiguous when you have text inputs, selects, checkboxes, and toggles
- Use the name everywhere: In design files, in code (class names, component names), in conversation, in documentation
Naming Test
A good pattern name passes three checks:
- A new team member can guess what it is from the name alone
- The name still works if the visual design changes completely
- The name works across all contexts where the pattern appears
Governance
Contribution Model
| Model | How It Works | Best For |
| Centralized | Dedicated system team owns everything. Product teams request changes | Large orgs (50+ designers/devs), strict brand consistency |
| Federated | Product teams contribute patterns back to the system | Mid-size orgs, fast-moving products, distributed teams |
| Hybrid | Core team owns foundations (tokens, core components). Product teams own domain-specific patterns | Most organizations — balances consistency with speed |
Pattern Lifecycle
Proposed → Under Review → Accepted → Built → Documented → Published → [Deprecated]Acceptance criteria for new patterns:
- Used (or needed) in 3+ contexts
- Clear, documented purpose
- Accessibility requirements met (WCAG 2.2 AA minimum)
- Responsive behavior defined
- Content guidelines included
- At least 2 team members reviewed
Keeping the System Alive
The #1 cause of design system failure is abandonment — enthusiasm at launch, decay within 6 months.
Prevention checklist:
- Dedicated owner (person or team) with allocated time — minimum 20% of one person's capacity
- System-first workflow: changes originate in the system, then propagate to products — never the reverse
- Automated sync between pattern library and production code
- Changelog published with every release
- Quarterly audit comparing library against live product
- Contribution process documented and frictionless — using the system must be easier than building from scratch
- Regular "state of the system" reviews with stakeholders
System Maturity Levels
| Level | Characteristics | Focus |
| 1 — Ad hoc | No shared patterns. Each feature built independently | Conduct interface inventory. Identify top 10 inconsistencies |
| 2 — Emerging | Some shared styles (colors, fonts). No component library | Define tokens. Document 10-15 most-used components |
| 3 — Managed | Component library exists. Adoption is partial | Governance process. Adoption metrics. Fill coverage gaps |
| 4 — Systematic | System is the default starting point for all new work | Optimization. Cross-platform support. Advanced theming |
| 5 — Embedded | System drives product decisions. Team culture centers on it | Innovation. Community contributions. Industry leadership |
Most organizations should target Level 3 within the first year and Level 4 within two years. Level 5 is rare and requires significant organizational commitment.
Common Mistakes
| Mistake | Why It Fails | Instead |
| Building the full system before any product uses it | No feedback loop. Patterns don't match real needs | Build alongside a real product. Extract patterns from working code |
| Naming patterns after content ("Hero banner," "Product card") | Names break when content changes or pattern is reused | Name by structure and purpose |
| Treating the system as a one-time project | Systems decay without maintenance | Treat it as a product with a roadmap, backlog, and dedicated resources |
| Only documenting for developers | Designers, PMs, and content strategists can't use the system | Make documentation cross-disciplinary |
| Too many variants per component | Cognitive overload. Inconsistent usage | Max 5 variants. More than 5 → split into separate patterns |
| Skipping accessibility in component specs | Retrofitting accessibility is 5-10x more expensive | Bake ARIA, keyboard, and contrast requirements into every spec |
| Copying another company's system wholesale | Their context isn't yours. Their trade-offs aren't yours | Learn from others. Build for your specific product, team, and users |
Reference Files
Load these for deeper guidance on specific topics:
references/component-hierarchy.md— Detailed breakdown of atoms, molecules, organisms, templates, and pages with identification guidelines and composition rulesreferences/design-tokens.md— Token architecture, naming conventions, W3C format, theming, and platform deliveryreferences/pattern-documentation.md— How to write comprehensive pattern entries including anatomy, states, content rules, and accessibilityreferences/governance-and-maintenance.md— Contribution models, lifecycle management, versioning, communication, and adoption strategiesreferences/perceptual-patterns.md— Color systems, typography scales, spacing systems, motion principles, and voice & tone guidelines
Templates
templates/component-spec-template.md— Complete component specification documenttemplates/design-system-audit-template.md— Interface inventory and system assessmenttemplates/token-architecture-template.md— Design token structure and naming
Examples
examples/system-audit-walkthrough.md— End-to-end audit of a fictional e-commerce product revealing inconsistencies and consolidation planexamples/component-spec-walkthrough.md— Specifying a Card component from purpose through accessibility
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

