Agent skill

Usage Guidelines

by murphytrueman in murphytrueman/design-system-ops

Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card.

MITAuto-check passedFrontend & Design

Install Usage Guidelines

skills CLI
$ npx skills add murphytrueman/design-system-ops --skill usage-guidelines -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install murphytrueman/design-system-ops usage-guidelines --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/murphytrueman/design-system-ops.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/usage-guidelines .claude/skills/usage-guidelines && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
usage-guidelines
GitHub stars
203
Token cost
~3.6k tokens
SKILL.md length
1,868 words
Files
1
Skills in repo
36
Repo updated
First seen
Licence
MIT

At a glance

Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card.

  • Works in 7 steps: Read the component → Gather component information → Write the usage guidelines → …
  • Tasks that involve Accessibility
  • SKILL.md covers Before you begin: verify…, Context, Step 0: Read the component and Step 1: Gather component…, plus 6 more sections
  • Calls npx

What it does

Usage Guidelines is an agent skill from murphytrueman/design-system-ops. Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card. Triggers: usage guidelines for X, do's and don'ts for X. Choosing between components: component-decision-tree. Multi-component patterns: pattern-documentation.

Its SKILL.md is about 3.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Frontend & Design, covering Accessibility. The repository describes itself as: Claude Code skills for the work that keeps a design system alive. The licence is MIT.

When your agent uses it

  • Tasks that involve Accessibility

Example prompts

  • “s and don”
  • “/usage-guidelines”

Requirements

  • Node.js
  • Pre-approved tools (allowed-tools): Read, Write, Grep, Glob, Bash(cat:*), Bash(find:*), Bash(head:*), Bash(ls:*)

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Read the component
  2. Gather component information
  3. Write the usage guidelines
  4. Review pass
  5. Generate the quick-reference card
  6. Voice and tone
  7. Summarise in chat

What it can do on your machine

Read from SKILL.md and the folder at commit f167898. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Grep
    • Glob
    • Bash(cat:*)
    • Bash(find:*)
    • Bash(head:*)
    • Bash(ls:*)

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • npx

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use npx, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Usage Guidelines loads about 3.6k tokens when it runs. Until then it costs about 76 tokens; SKILL.md has 1,868 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~76
When it runs · the whole SKILL.md, loaded when a task matches
~3.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from murphytrueman/design-system-ops at commit f167898, republished under its MIT licence (© murphytrueman). 1,868 words, ~3,606 tokens.

Download SKILL.mdSave it as .claude/skills/usage-guidelines/SKILL.md (or your agent's skills folder).
name
usage-guidelines
description
Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card. Triggers: usage guidelines for X, do's and don'ts for X. Choosing between components: component-decision-tree. Multi-component patterns: pattern-documentation.
allowed-tools
Read, Write, Grep, Glob, Bash(cat:*), Bash(find:*), Bash(head:*), Bash(ls:*)
references
../../knowledge-notes/ai-readiness.md, ../../knowledge-notes/component-bestiary-reference.md, ../../knowledge-notes/output-discipline.md

Usage guidelines

A skill for writing component usage guidelines that cover the full usage contract: when to use, when not to, edge cases, anti-patterns, and accessibility guidance integrated throughout. Output reads as guidance a designer or developer can act on immediately, not a style guide entry that restates what is already visible in the component.

Before you begin: verify references

Confirm that every path in this skill's frontmatter references: exists relative to this SKILL.md. If any is missing, stop: the install is incomplete, usually because a flattening installer (for example npx skills install) dropped the repo-root knowledge-notes/ directory. Tell the user to reinstall by a method in 1-INSTALL.md and run verify-install.sh from the install root. Proceed without the references only if the user explicitly says to, and then say in the output that it was produced without the pack's reference material.

Context

Most component usage guidelines have the same structural problem: they describe the component rather than guiding its use. "The button component is used to trigger actions" is a description. "Use a primary button for the single most important action in a given context — never more than one per view" is guidance. The first tells you what exists. The second tells you how to use it correctly.

The goal here is the second kind. Guidelines that are worth writing are guidelines that would prevent a real mistake someone on a consuming team could plausibly make.


Step 0: Read the component

Before asking the user anything, read what's available:

  • .ai/metadata/<Component>.metadata.json, if metadata-schema-generator has run: props, states and the accessibility contract, each with a provenance marker. Take them from there and don't re-derive them; carry the markers through
  • the component's source (variants, rendered element, ARIA attributes, key handlers, focus calls), its stories, existing docs, and the Figma component when a Figma MCP is connected
  • the system's voice and tone guide: system.content_guidelines_path in config, or CONTENT.md, VOICE.md, writing-guidelines.md or similar, so the content guidelines section uses the team's rules from the start rather than generic UX writing advice

Much of Step 1 is answered there. If the component can't be found in source, stories, docs, metadata or Figma, stop and ask where it lives; guidelines for a component you haven't seen are generic by construction.

Provenance rule. Every behaviour the guidelines state as fact (variants, keyboard interaction, focus, announcements, contrast) traces to source, stories, docs, Figma or the user. Anything you can't confirm is marked "unverified"; misuse you haven't seen evidence of is marked "anticipated". See "Every figure and fact needs a source" in the output-discipline knowledge note.

Step 1: Gather component information

Confirm from Step 0, and ask the user for the rest:

  • Component name and the system it belongs to
  • Available variants or configurations
  • Any existing documentation to build on or replace
  • Known misuse patterns the team has actually seen in production — these are the most valuable input
  • Any accessibility requirements already established for the component

The known misuse patterns are critical. Guidelines written from abstract principle tend to address imaginary mistakes. Guidelines written from observed patterns address real ones.

Step 2: Write the usage guidelines


[Component name] usage guidelines

Version: [design system version: ask; don't infer] Last updated: [date]


Overview

One to two sentences. What does this component do and what user need does it serve? Write this as the answer to "why does this component exist" not "what does it look like."


When to use

Write as specific conditions, not general descriptions. Each condition should be concrete enough that a designer could read it and make a decision.

Cover the primary use case first, then secondary use cases. Three to five conditions is usually the right scope — more than that and the guidelines are covering for an unclear component contract.

Format: "Use [component name] when [specific condition]."

Examples (illustrative Button):

  • Use a primary button for the single most important action in a context. There should be at most one primary button in any given view.
  • Use a secondary button for an action that is available but not the expected next step. Secondary buttons often appear alongside primary buttons to give users an alternative.
  • Use a ghost button when the action is available but should not visually compete with other actions or content on the page.

When not to use

As important as the above, and often more valuable. Each entry should name a specific misuse and point to an alternative.

Format: "Do not use [component name] for [misuse condition]. Use [alternative] instead."

Examples (illustrative Button):

  • Do not use a button for navigation to another page. Use a link. Buttons trigger actions; links navigate. Using a button for navigation misrepresents the interaction to screen readers and keyboard users.
  • Do not use more than one primary button in the same context. If two actions feel equally important, reconsider the information hierarchy.
  • Do not use a button when no action occurs. If the element is decorative or informational, it is not a button.

Variants and configurations

For each variant or major configuration option: one sentence on what it is for and one sentence on when to use it. Do not repeat information already in the component API — this section should add intent context, not restate prop values.

Only document variants that require usage judgment. If a variant is self-explanatory ("size" on a component that comes in sm, md, and lg) skip it or document it briefly. Spend the space on variants where misuse is plausible.


Edge cases

Edge cases are the situations the happy path documentation does not cover. They are the most important section for preventing real-world mistakes and the most commonly omitted.

Document:

  • What happens when the label text is very long?
  • What happens in a right-to-left layout?
  • What happens when the component is used on a non-white or non-standard background?
  • What happens when multiple instances appear in close proximity?
  • What happens when the action is destructive and irreversible?

Not every component has every type of edge case. Only document the edge cases that are real for this component — do not produce a generic list.


Accessibility

Do not relegate accessibility to a separate section or an afterthought. For each point in the usage guidelines where an accessibility concern is relevant, integrate it in context.

Additionally, provide a consolidated accessibility reference covering the items below. Each line ends with its source in brackets: a file and line, a story id, a metadata provenance marker, a computed ratio, a Figma node, or user. Anything with no source is marked "unverified" rather than described from what components like this usually do.

  • Keyboard interaction: which keys, in which order, with which outcomes
  • Focus behaviour: where focus sits in default state, how it changes on interaction
  • Screen reader: what gets announced, when, and how that announcement is produced
  • Minimum touch target: if relevant, state the minimum size and how the component handles it
  • Colour contrast: which colour combinations have been checked, by what, and against which level. The baseline is WCAG 2.2 AA. Some legal baselines (e.g. EN 301 549) still reference WCAG 2.1 AA; use that if it's the team's obligation.

Accessibility guidance should be specific to this component. Do not cite WCAG criteria as the guidance itself — say what the component does.


Show full SKILL.md (668 more words)Show less
Anti-patterns

Name the three to five most common ways this component is misused, each as a clear statement with a reason and a correction.

Write these based on observed misuse patterns where possible. Anti-patterns derived from production experience are consistently more useful than anti-patterns derived from abstract reasoning. Where you have no evidence of the misuse (from the user, reviews, drift findings or code), mark the anti-pattern "anticipated" in its heading.

Anti-pattern template:

For each anti-pattern, use this consistent structure to ensure they are actionable:

**Anti-pattern: [Short name]**
What happens: [One sentence describing the misuse]
Why it's harmful: [One sentence on the specific consequence — accessibility, consistency, UX, or maintenance]
What to do instead: [One sentence with the correct approach]
How to detect: [One sentence on how to spot this in a review — what to look for in code, design, or Storybook]

Example (illustrative):

**Anti-pattern: Navigation button**
What happens: A Button component is used to navigate to another page.
Why it's harmful: Screen readers announce it as a button, not a link — users expect an action, not navigation. Keyboard behaviour differs (buttons activate on Space, links do not).
What to do instead: Use a Link component styled to match the desired visual weight.
How to detect: Look for Button components with onClick handlers that call router.push(), window.location, or equivalent navigation functions.

The "how to detect" field is particularly valuable for code reviewers and linting rules — it translates the anti-pattern from a principle into a checkable condition.


Content guidelines

If the component displays text that product teams write (button labels, error messages, empty state copy, tooltip content): include brief content guidelines covering the appropriate tone, length, and framing.

These are particularly important for:

  • Buttons: action-oriented labels, verb-led, specific
  • Error messages: cause and resolution, not just notification
  • Empty states: context-appropriate next action, not generic "no data found"
  • Tooltips: supplementary, not required reading

If content guidelines are not relevant to this component, skip this section.


Cross-references to components that are commonly confused with this one, or commonly used alongside it. For each:

  • Component name
  • One sentence distinguishing it from this component, or describing how they work together

Step 3: Review pass

Read the draft once for two things: accessibility appears in context (edge cases, anti-patterns and "when not to use" each carry it, not only the dedicated section), and every conditional is specific enough for a developer to implement, not only for a designer to recognise.

Step 5: Generate the quick-reference card

Full guidelines are valuable for deep understanding. But in a code review, a design crit, or a sprint, teams need a one-page reference they can check in 30 seconds. After writing the full guidelines, generate a condensed quick-reference card.

Quick-reference card format:

markdown
## [Component name] — Quick reference

**Use when:** [3–5 bullet points, one line each, from "When to use"]

**Don't use when:** [3–5 bullet points, one line each, from "When not to use"]

**Watch for:**
- [Anti-pattern 1 — one line]
- [Anti-pattern 2 — one line]
- [Anti-pattern 3 — one line]

**Accessibility:** [keyboard pattern] | [required ARIA] | [focus behaviour — one line]

**Related:** [Component A] for [distinction] | [Component B] for [distinction]

The quick-reference card should fit in roughly 150 words. It is a lookup tool, not a learning document. Every line should be a decision aid — if it does not help someone make a choice in the moment, it does not belong on the card.

Deliver both the full guidelines and the quick-reference card as separate sections in the output. Teams can publish the quick-reference card alongside the component in their documentation site for fast access.


Step 6: Voice and tone

The content guidelines section uses the system's own voice and tone guide, found in Step 0. Quote its rules (sentence case, verb-led labels, error message shape) with a component-specific example each. If no guide exists, use general UX writing principles and add one line: "These content guidelines use general principles; document the system's voice and tone and reference it here."

Step 7: Summarise in chat

End with a short chat summary:

  • Headline: the component and what the guidelines cover
  • Written: file path if saved, otherwise "in chat"
  • Marked: each "unverified" accessibility claim and each "anticipated" anti-pattern, so the team knows what to confirm
  • Scope: the block from the output-discipline knowledge note, naming the source, stories, docs and Figma actually read

Quality checks

  • "When to use" conditions are specific enough to make a decision from
  • "When not to use" entries each name an alternative
  • Edge cases are real for this component, not generic
  • Accessibility is integrated throughout, not siloed at the end
  • Anti-patterns are derived from observed misuse, or clearly noted as anticipated misuse if observed examples are not available
  • Content guidelines are included for text-bearing components and use the system's own voice/tone when documented, not just generic UX writing advice
  • Keyboard, focus, announcement and contrast claims each cite their source in brackets; the rest is marked "unverified"
  • Where .ai/metadata/ exists, props and the accessibility contract came from it with their provenance markers
  • Guidelines work for both designers and developers
  • Nothing in the guidelines restates what is already visible in the component — every line adds usage judgment, not description
  • A quick-reference card (~150 words) is included alongside the full guidelines

© murphytrueman, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/usage-guidelines of murphytrueman/design-system-ops.

Open the folder on GitHubat commit f167898

Compare with similar skills

Usage Guidelines next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Usage Guidelines compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Usage Guidelines this skillmurphytrueman/design-system-ops203—~3.6kAutomated safety check: PassMIT
Web Interface Guidelines Reviewervercel-labs/openreview1.7k98 repos~308Automated safety check: PassNone
Accessibility Reviewmarkmead/hyperui12k1 repos~1.1kAutomated safety check: PassMIT
Web Animation DesignbaptisteArno/typebot.io11k2 repos~2.7kAutomated safety check: PassCustom licence
Accessibility Fixeribelick/ui-skills9.5k4 repos~1.2kAutomated safety check: PassMIT
Wcag Audit PatternsvmDeshpande/ai-agent-automation17811 repos~610Automated safety check: PassApache-2.0

Similar skills

  • Web Interface Guidelines Reviewer

    vercel-labs/openreview

    Official

    Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my…

    1.7k GitHub starsUsed in 98 repos~308 tokens
    Frontend & DesignAuto-check passed
  • Accessibility Review

    markmead/hyperui

    Run a WCAG 2.1 AA accessibility audit on a design or page. An agent skill from markmead/hyperui.

    12k GitHub starsUsed in 1 repo~1.1k tokens
    Frontend & DesignAuto-check passed
  • Web Animation Design

    baptisteArno/typebot.io

    Guides easing, timing and animation choices for UI motion, based on a web animation course, and reviews existing animations in a before-and-after table.

    11k GitHub starsUsed in 2 repos~2.7k tokens
    Frontend & DesignAuto-check passed
  • Accessibility Fixer

    ibelick/ui-skills

    Audits and fixes HTML accessibility problems such as ARIA labels, keyboard navigation, focus management, contrast and form errors with minimal changes.

    9.5k GitHub starsUsed in 4 repos~1.2k tokens
    Frontend & DesignAuto-check passed
  • Wcag Audit Patterns

    vmDeshpande/ai-agent-automation

    Conduct WCAG 2.2 accessibility audits with automated testing, manual verification, and remediation guidance.

    178 GitHub starsUsed in 11 repos~610 tokens
    Frontend & DesignAuto-check passed
  • Baseline UI

    ibelick/ui-skills

    Applies a fixed set of UI rules for stack, components, interaction, animation, typography and layout, or reviews a file against them with concrete fixes.

    9.5k GitHub starsUsed in 8 repos~855 tokens
    Frontend & DesignAuto-check passed

More from murphytrueman/design-system-ops

All 36 skills in this repo
  • Agent Instructions

    murphytrueman/design-system-ops

    Write the AGENTS.md that tells coding agents how to use this design system: where things live, sourced rules, how to check work, what not to do; Claude, Cursor or Copilot pointers on request.

    203 GitHub stars~2.3k tokensUpdated 14 days ago
    Auto-check passed
  • AI Component Description

    murphytrueman/design-system-ops

    Write a six-section prose description (purpose, props, anti-patterns, composition, accessibility, examples) for a Figma component's description field so LLMs read it via MCP.

    203 GitHub stars~4.7k tokensUpdated 14 days ago
    Auto-check passed
  • Change Communication

    murphytrueman/design-system-ops

    Write release notes, a migration guide and a team announcement for a design system change that is already decided, scaled to its impact.

    203 GitHub stars~3.4k tokensUpdated 14 days ago
    Auto-check passed
  • Codebase Index

    murphytrueman/design-system-ops

    Generate machine-readable index files in .ai/index/ (component inventory, uses/usedBy graph, stats) for AI agents.

    203 GitHub stars~4.7k tokensUpdated 14 days ago
    Auto-check passed
  • Codemod Generator

    murphytrueman/design-system-ops

    Generate tested jscodeshift/postcss codemods for design system migrations: token renames, prop renames or removals, import paths, component swaps.

    203 GitHub stars~4.9k tokensUpdated 14 days ago
    Auto-check passed
  • Component API Validator

    murphytrueman/design-system-ops

    Audit prop APIs across a component library: naming consistency, boolean/default patterns, type coverage, exported types, breaking changes between versions.

    203 GitHub stars~4.3k tokensUpdated 14 days ago
    Auto-check passed

Questions about Usage Guidelines

What does Usage Guidelines do?

Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card. Usage Guidelines is an agent skill from murphytrueman/design-system-ops. Write usage guidelines for one named component: when to use, when not to, edge cases, anti-patterns, a11y, quick-reference card.

When should I use Usage Guidelines?

Usage Guidelines fits situations like: tasks that involve Accessibility.

How do I install Usage Guidelines in Claude Code?

Run `npx skills add murphytrueman/design-system-ops --skill usage-guidelines -a claude-code`. Or copy the skill folder (skills/usage-guidelines in murphytrueman/design-system-ops) into .claude/skills/usage-guidelines in your project. Claude Code loads it when a task matches its description.

How do I install Usage Guidelines in Codex?

Run `npx skills add murphytrueman/design-system-ops --skill usage-guidelines -a codex`. Or copy the skill folder (skills/usage-guidelines in murphytrueman/design-system-ops) into .agents/skills/usage-guidelines in your project. Codex loads it when a task matches its description.

Can I use Usage Guidelines in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add murphytrueman/design-system-ops --skill usage-guidelines -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/usage-guidelines, .gemini/skills/usage-guidelines, .github/skills/usage-guidelines and .opencode/skills/usage-guidelines in your project.

What does Usage Guidelines need to run?

Going by SKILL.md and its folder, Usage Guidelines needs the command-line tools its instructions call (npx). Our summary lists: Node.js. Its frontmatter pre-approves these tools: Read, Write, Grep, Glob, Bash(cat:*), Bash(find:*), Bash(head:*), Bash(ls:*).

Does Usage Guidelines access the network?

SKILL.md contains no URLs. Its commands use npx, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Usage Guidelines safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Usage Guidelines use?

Usage Guidelines is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Usage Guidelines use?

About 3.6k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Usage Guidelines?

Skills that share tags, products or a category with Usage Guidelines: Web Interface Guidelines Reviewer (vercel-labs/openreview, 1.7k stars), Accessibility Review (markmead/hyperui, 12k stars), Web Animation Design (baptisteArno/typebot.io, 11k stars) and Accessibility Fixer (ibelick/ui-skills, 9.5k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Usage Guidelines?

murphytrueman (a GitHub user) maintains it in murphytrueman/design-system-ops, which has 203 GitHub stars. The repository holds 36 skills in this directory. The repository was last updated on September 24, 2026.

Source: murphytrueman/design-system-ops on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.