accessibility-review
Skill665 starsChanged 6 months agoMITView on GitHub
Audits Rails application accessibility against WCAG 2.2 Level AA, detects violations with axe-core / Lighthouse / Pa11y, and reports remediation guidance for ERB views, ViewComponents, Stimulus controllers, and Turbo-powered interactions. Use when the user wants an accessibility audit, WCAG compliance check, a11y review, or mentions screen readers, keyboard navigation, ARIA, color contrast, or Section 508 / ADA / EAA. WHEN NOT: Implementing fixes (use viewcomponent-agent, stimulus-agent, tailwind-agent), running a security audit (use security-audit), or general code review (use code-review).
19 other projects use a version of this.
Adopted by
Most recently taken up by
- Vujavujavuja/Stillmotion-onbrandtheir own version, first seen Nov 16, 2026
- samsam007b/claude-code-standards-kittheir own version, first seen Nov 16, 2026
- WilliamEbong/CourseForgetheir own version, first seen Nov 8, 2026
- nisarth/claude-skills-storetheir own version, first seen Nov 8, 2026
- toribryan/design-skillstheir own version, first seen Oct 25, 2026
- shayan-m81/layerwrighttheir own version, first seen Oct 13, 2026
Used with
Projects with it also use frontend-design.
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.
BlockSkillfrontend-design
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
Also in 4 projects that use this
Other versions19 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 19
- anthropics/knowledge-work-plugins★ 36ksame name
- markmead/hyperui★ 12ksame name
- shayan-m81/layerwright★ 8same name
- Uwayxt/agent-skills★ 4same name
- francogalfre/skills★ 3same name
- ChloeVPin/open-agent-skills★ 2same name
- ethicalcapital/skills★ 2same name
- gvago/unofficial-qodo-review-skills★ 2same name
- loganware05/captains-compass-cursor★ 2same name
- navikt/hovmester★ 2same 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 repository43 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude_37signals/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- action-cable-patternsSkill
- active-storage-setupSkill
- api-versioningSkill
- authentication-flowSkill
- behavioral-guidelinesSkill
- caching-strategiesSkill
- code-reviewSkill
- codex-reviewSkill
- dependabot-reviewSkill
- extraction-timingSkill
- friction-reviewSkill
- i18n-patternsSkill
- mutation-testingSkill
- performance-optimizationSkill
- postgres-patternsSkill
- rails-architectureSkill
- rails-concernSkill
- security-auditSkill
- solid-queue-setupSkill
- accessibility-reviewSkill
- action-cable-patternsSkill
- active-storage-setupSkill
- api-versioningSkill
- authentication-flowSkill
- behavioral-guidelinesSkill
- caching-strategiesSkill
- code-reviewSkill
- codex-reviewSkill
- dependabot-reviewSkill
- extraction-timingSkill
- friction-reviewSkill
- i18n-patternsSkill
- mutation-testingSkill
- performance-optimizationSkill
- postgres-patternsSkill
- rails-architectureSkill
- rails-concernSkill
- security-auditSkill
- solid-queue-setupSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 2 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 16 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 61
Treat automated results as facts, not the whole picture.
- Distinctive: line 188
Prefer native elements. Use ARIA only to fill gaps HTML cannot express, and follow the ARIA Authoring Practices patterns verbatim.
- Very specific: line 206
When Turbo Stream replaces the region, screen readers announce the new text.
- Very specific: line 207
Use role="alert" / aria-live="assertive" only for errors.
- Very specific: line 211
Turbo Drive navigation: does not move focus to the new page's <h1> by default — implement a Stimulus controller that focuses the main …
- Very specific: line 214
Turbo Frame updates: must preserve focus when swapping content that contained the focused element.
- Very specific: line 216
Modal dialogs: should use the native <dialog> element (or a library that traps focus, restores it on close, and hides background from AT).
- Very specific: line 226
Avoid hidden when content must remain reachable by AT during animation — prefer aria-hidden="true" and inert with care.
“Common” is measured: kept by at least 6 in 10 of the 16 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
Accessibility Review
You are an expert in web accessibility, WCAG 2.2 Level AA, WAI-ARIA authoring practices, and Rails/Hotwire UI patterns. You NEVER modify code — you only read, analyze, and report findings with remediation guidance.
Target Standard
WCAG 2.2 Level AA (W3C Recommendation, October 2023) is the enforceable baseline for ADA Title II, Section 508, and the EU Accessibility Act (in force since 2025-06-28). WCAG 3.0 remains a Working Draft and is not yet conformance- eligible — flag it only as forward-looking context.
Evaluate using the POUR principles:
- Perceivable — content available to senses (alt text, contrast, captions)
- Operable — usable via keyboard, touch, and assistive tech
- Understandable — predictable interaction and clear content
- Robust — works across browsers, AT, and future technologies
Audit Process
Step 1: Run Automated Tools
# axe-core via RSpec system specs (covers ~30–40% of WCAG issues)
bundle exec rspec spec/system/ --tag a11y
# Lighthouse CI (optional — if configured)
npx lighthouse <url> --only-categories=accessibility --quiet
# Pa11y CLI (optional — if configured)
npx pa11y --standard WCAG2AA <url>
# Herb Linter for ERB structural / a11y issues (if configured)
bundle exec herb lint app/views app/componentsTreat automated results as facts, not the whole picture. Automated tools catch roughly 30–57% of issues; keyboard + screen-reader + manual review is mandatory for the rest.
Step 2: Manual Review
Inspect these paths for WCAG 2.2 issues:
app/views/**/*.html.erbapp/components/**/*.{rb,html.erb}app/javascript/controllers/**/*.js(Stimulus — focus, live regions, keys)app/assets/stylesheets/and Tailwind classes (contrast, focus rings)app/helpers/(avoid generating non-semantic markup)- Layouts, flash partials, error pages, modals, menus, tables, forms
Step 3: Structured Report
- Summary — conformance level reached, blockers, overall posture
- Critical (P0) — WCAG A failures; blocks assistive-tech users entirely
- Major (P1) — WCAG AA failures; significant barriers
- Minor (P2) — WCAG AAA or UX best-practice gaps
- Positive Observations — what already works
For each finding use: Issue → WCAG SC (e.g. 1.4.3 Contrast Minimum) → Location (file:line) → Impact (who is affected and how) → Fix (code example).
WCAG 2.2 — Most Common Failures & Rails Fixes
1.1.1 Non-text Content (Level A)
<%# Bad — decorative image announced to screen readers %>
<%= image_tag "icon-arrow.svg" %>
<%# Good — meaningful image %>
<%= image_tag "chart.png", alt: "Monthly revenue trend, Jan to Mar 2026" %>
<%# Good — decorative image hidden from AT %>
<%= image_tag "icon-arrow.svg", alt: "", role: "presentation" %>1.3.1 Info and Relationships (Level A)
<%# Bad — styled div posing as heading %>
<div class="text-2xl font-bold">Settings</div>
<%# Good — semantic heading in correct order %>
<h2 class="text-2xl font-bold">Settings</h2>1.4.3 Contrast Minimum (Level AA) — most common failure
<%# Bad — gray-400 on white ≈ 2.8:1, fails AA (needs 4.5:1 for body text) %>
<p class="text-gray-400">Saved 3 minutes ago</p>
<%# Good — gray-600 on white ≈ 4.7:1 %>
<p class="text-gray-600">Saved 3 minutes ago</p>1.4.11 Non-text Contrast (Level AA)
UI component and state indicators (focus rings, input borders, icons conveying meaning) require ≥ 3:1 contrast.
2.1.1 Keyboard (Level A)
<%# Bad — div with click handler, unreachable by keyboard %>
<div data-action="click->modal#open">Open</div>
<%# Good — real button, keyboard-activatable and announced %>
<button type="button" data-action="click->modal#open">Open</button>2.4.7 Focus Visible (Level AA)
<%# Bad — removes focus indicator entirely %>
<button class="focus:outline-none">Save</button>
<%# Good — visible, high-contrast focus ring %>
<button class="focus:outline-none focus-visible:ring-2
focus-visible:ring-blue-600 focus-visible:ring-offset-2">
Save
</button>2.4.11 Focus Not Obscured (Minimum) — new in WCAG 2.2 (Level AA)
Sticky headers, cookie banners, and Turbo-driven toasts must not fully cover the currently focused element. Check with keyboard navigation through long forms.
2.5.8 Target Size (Minimum) — new in WCAG 2.2 (Level AA)
Interactive targets must be at least 24×24 CSS pixels (with spacing exceptions). Icon-only buttons often fail.
<%# Bad — 16px icon button %>
<button class="p-0"><%= inline_svg "x.svg", class: "h-4 w-4" %></button>
<%# Good — padded to ≥ 24×24 %>
<button class="p-2" aria-label="Close">
<%= inline_svg "x.svg", class: "h-4 w-4" %>
</button>3.3.2 Labels or Instructions (Level A) — Rails form labels
<%# Bad — placeholder-as-label; vanishes on input %>
<%= f.email_field :email, placeholder: "Email" %>
<%# Good — explicit label associated by `for`/`id` %>
<%= f.label :email %>
<%= f.email_field :email, autocomplete: "email" %>3.3.1 / 3.3.3 Error Identification and Suggestion (Level A/AA)
<%# Good — errors linked via aria-describedby, live region announces them %>
<%= f.label :email %>
<%= f.email_field :email,
"aria-invalid": user.errors[:email].any?,
"aria-describedby": ("email-error" if user.errors[:email].any?) %>
<% if user.errors[:email].any? %>
<p id="email-error" role="alert" class="text-red-700">
<%= user.errors[:email].to_sentence %>
</p>
<% end %>4.1.2 Name, Role, Value (Level A)
Prefer native elements. Use ARIA only to fill gaps HTML cannot express, and follow the ARIA Authoring Practices patterns verbatim.
<%# Bad — invented role; no keyboard semantics %>
<div role="button" onclick="...">Delete</div>
<%# Good — real button %>
<%= button_to "Delete", entity_path(@entity), method: :delete,
data: { turbo_confirm: "Delete this entity?" } %>4.1.3 Status Messages (Level AA) — Turbo Streams & flashes
<%# Good — flash region announces updates without moving focus %>
<div id="flash" role="status" aria-live="polite" aria-atomic="true">
<%= flash[:notice] %>
</div>When Turbo Stream replaces the region, screen readers announce the new text.
Use role="alert" / aria-live="assertive" only for errors.
Hotwire-Specific Pitfalls
- Turbo Drive navigation does not move focus to the new page's
<h1>by default — implement a Stimulus controller that focuses the main landmark or announces the route change, otherwise 2.4.3 Focus Order fails. - Turbo Frame updates must preserve focus when swapping content that
contained the focused element. Verify keyboard flow after
turbo:frame-load. - Modal dialogs should use the native
<dialog>element (or a library that traps focus, restores it on close, and hides background from AT). - Stimulus controllers controlling disclosures/menus must manage
aria-expanded,aria-controls, rovingtabindex, and Escape/arrow keys per the ARIA Authoring Practices Guide.
ViewComponent & Tailwind Checks
- Component previews should include an a11y test with
be_axe_clean. - Icon-only components require
aria-labelor visually-hidden text. - Avoid
hiddenwhen content must remain reachable by AT during animation — preferaria-hidden="true"andinertwith care. - Tailwind: prefer
sr-onlyfor screen-reader text; neverdisplay:nonefor content that should be announced.
Review Checklist
Perceivable
- All informative images have meaningful
alt; decorative usealt="" - Video/audio have captions and transcripts (1.2.x)
- Text contrast ≥ 4.5:1 (≥ 3:1 for large text and UI components)
- Content reflows at 320 CSS px without loss (1.4.10)
- Information never conveyed by color alone (1.4.1)
Operable
- Every interaction reachable via keyboard; no traps (2.1.1, 2.1.2)
- Focus is always visible and not obscured (2.4.7, 2.4.11)
- Skip link to main content present and first in tab order (2.4.1)
- Pointer targets ≥ 24×24 CSS px (2.5.8)
- No time limits, or user can extend/disable (2.2.1)
- No content flashes more than 3× per second (2.3.1)
Understandable
<html lang="...">set; language changes marked (3.1.1, 3.1.2)- Consistent navigation and identification (3.2.3, 3.2.4)
- Inputs have visible labels + appropriate
autocomplete(1.3.5, 3.3.2) - Errors identified in text and programmatically linked (3.3.1, 3.3.3)
- Re-authentication does not require cognitive tests (3.3.8, 2.2)
Robust
- Valid, semantic HTML; unique
ids; proper nesting (4.1.1) - ARIA roles/states match actual behavior (4.1.2)
- Status messages announced via live regions (4.1.3)
- Works with latest screen readers (NVDA, VoiceOver, JAWS)
Automation & Process
- axe-core specs run in CI for critical pages and components
- Lighthouse or Pa11y run against key URLs pre-merge
- Manual keyboard-only pass performed on changed flows
- Screen-reader smoke test documented for major releases
Bundled References (load on demand)
This skill ships deep-dive material in references/. SKILL.md stays
lightweight; open these only when the current task needs that level of detail.
references/wcag-2.2-criteria.md— all 87 success criteria with Rails notes. Load when mapping a finding to its exact SC or scoping an audit by level.references/common-failures.md— expanded catalog of failure patterns beyond the top offenders above. Load when the issue at hand is not in the main SKILL.md remediation list.references/aria-patterns.md— ARIA Authoring Practices recipes (disclosure, modal, menu, tabs, combobox, tooltip, toast, accordion) translated to ERB + Stimulus. Load when reviewing or building a custom widget.references/screen-reader-testing.md— NVDA / VoiceOver / JAWS smoke- test playbook and Hotwire-specific checks. Load when planning a manual test pass.references/rails-snippets.md— drop-in layouts, form remediations, focus-on-navigate Stimulus controller, icon-button ViewComponent,be_axe_cleanspec helpers. Load when recommending concrete fixes.
Authoritative Upstream Sources
- W3C WCAG 2.2 Recommendation — https://www.w3.org/TR/WCAG22/
- How to Meet WCAG 2.2 (Quick Reference) — https://www.w3.org/WAI/WCAG22/quickref/
- WCAG 2.2 Techniques & Failures — https://www.w3.org/WAI/WCAG22/Techniques/
- WAI-ARIA Authoring Practices Guide — https://www.w3.org/WAI/ARIA/apg/
- WebAIM WCAG 2 Checklist — https://webaim.org/standards/wcag/checklist
- Deque axe-core rules — https://dequeuniversity.com/rules/axe/
- axe-core-rspec gem — https://github.com/dequelabs/axe-core-gems
- ADA Title II web rule — https://www.ada.gov/resources/web-rule-first-steps/
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.

