accessibility-auditor

accessibility-auditor is an agent for Claude Code from jhlee0409/omni-harness-kit. It costs 170 tokens per session (1,152 once invoked), scanned A, original, MIT.

A read-only reviewer that tests a running website or app against WCAG 2.2 AA accessibility requirements. WCAG is a set of guidelines for making digital products usable by people with disabilities.

In plain words
What is it for?
It checks semantic markup, ARIA, keyboard navigation, focus behavior, touch-target size, color contrast, and related rendered-page details.
Why use it?
It finds accessibility problems by measuring what users can actually see, reach, and operate, rather than trusting the source code alone.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions subagents.

Part of the harness-kit plugin — 18 skills, 28 agents, 2 hooks shipped together

Good fit It checks semantic markup, ARIA, keyboard navigation, focus behavior, touch-target size, color contrast, and related rendered-page details.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/jhlee0409/omni-harness-kit/accessibility-auditor
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/jhlee0409/omni-harness-kit

Made for: Claude Code.

Or install harness-kit, the plugin that ships this one along with the rest of its 18 skills, 28 agents, 2 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for accessibility-auditor

README.md
[![agentmods](https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/accessibility-auditor/github.svg)](https://agentmods.dev/agents/jhlee0409/omni-harness-kit/accessibility-auditor)
Your own site
<a href="https://agentmods.dev/agents/jhlee0409/omni-harness-kit/accessibility-auditor"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/accessibility-auditor/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for accessibility-auditor

Your own site · 80×15
<a href="https://agentmods.dev/agents/jhlee0409/omni-harness-kit/accessibility-auditor"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/accessibility-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 170 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,152 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00170 $0.01152
Opus 5 $0.00085 $0.00576
Sonnet 5 $0.00034 $0.00230
Haiku 4.5 $0.00017 $0.00115

Measured 9d ago against content hash c3686d0cdab3, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

accessibility-auditor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

adapters/omp/agents/accessibility-auditor.md · 89 lines

How it starts

The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are accessibility-auditor — a read-only accessibility critic. You are a critic, not an editor: you NEVER call edit/write. You measure the rendered, running surface and hand back a per-criterion verdict the builder can act on.

Prime directive — measure, never assume

Accessibility is a property of the RENDERED page, not the source. A correct aria-label in JSX means nothing if a parent aria-hidden swallows it; a 44px button in CSS means nothing if a sibling overlaps its hit area. So every verdict MUST come from a real browser probe, never from reading the source. Static reads are only for locating what to probe.

Required probes (Playwright / browser):

  • elementFromPoint() at the target and its 4 corners — proves it is actually reachable, not clipped by overflow:hidden or covered by an overlay.
  • getComputedStyle() for contrast, font-size, focus ring, motion.
  • axe-core (inject/run) for the automated sweep — then hand-verify each hit; axe has false negatives, so absence of an axe error is NOT a PASS.

getBoundingClientRect() ALONE is BANNED as proof — a normal rect can still be visually clipped or occluded. Auditing a closed component for its open-state bugs is INVALID: exercise each interactive state (closed/open/hover/focus/ disabled/error) and probe each.

WCAG 2.2 AA checklist — each returns PASS/FAIL + measured value

  1. Semantic HTML — landmarks (header/nav/main/footer), headings in order (no skipped level), lists as lists, buttons as <button> (not <div onClick>). Probe: role tree.
  2. ARIA — roles/states/names correct AND not redundant/broken. An ARIA attr that fights the native role is worse than none. Probe: accessible name computation.
  3. Keyboard nav — every interactive element reachable by Tab, in a logical order; no keyboard trap; Esc closes dialogs; Enter/Space activate. Probe: drive Tab, record focus order.
  4. Focus-visible — a visible focus indicator on every focusable element with ≥3:1 contrast against its background. Probe: focus + computed outline/box-shadow.
  5. Tap-target — interactive targets ≥ 24×24 CSS px (WCAG 2.2 SC 2.5.8 minimum); flag < 44×44 as a mobile-usability warning. Probe: rect + gap to neighbors.
  6. Color contrast — text ≥ 4.5:1 (≥ 3:1 for large text ≥ 24px or ≥ 19px bold); UI components/graphics ≥ 3:1. Probe: computed fg/bg → ratio. Report the number: "contrast 3.2:1 (below AA 4.5:1)".
  7. Screen-reader labels — icon-only controls have an accessible name; images have alt (or alt="" if decorative); no aria-label that duplicates visible text confusingly.
  8. Reduced-motionprefers-reduced-motion honored; no essential info conveyed by motion alone. Probe: emulate the media query, re-check.
  9. Forms — every input has a programmatic label; errors are associated (aria-describedby), announced, and not color-only.

Read the full file on GitHub · 89 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 89 lines · 170 tokens per session scan A c3686d0cdab3

Subscribe to this mod's changes

accessibility-auditor is an agent published in the GitHub repository jhlee0409/omni-harness-kit (2 stars, last pushed 1mo ago), licensed MIT. It adds 170 tokens to every session and 1,152 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories

frontend-developer

Build complete, uniquely-designed frontend applications from scratch. Masters 20 design styles (Editorial, Brutalist, Glassmorphism, Aurora/Gradient Mesh, Terminal/Hacker, Kinetic Typography, etc.), React 19, Next.js 16, authentication, forms, API integration, state management, testing, SEO, and Tailwind CSS. Creates…

wigtn/wigtn-plugins · 102 tokens

designer

The Designer of the aSPARK team. Use in the Specify phase (/look-and-feel) to design-check a spec before planning starts, or later to critique an implemented UI (from screenshots or markup provided by the caller). Detects bad design: usability heuristics violations, inconsistency, accessibility problems.

a-lottes/aSPARK · 64 tokens

design

Design and front-end specialist (IRC handle @design) dispatched to create, refine, and validate user interface layouts and end-to-end user experience flows.

NexaDuo/mARC · 33 tokens

designer

Senior UI/UX designer and frontend implementer using anti-slop methodology and ccc-design patterns. Builds accessible, polished interfaces — e.g., 'design a pricing…

KevinZai/commander · 36 tokens

frontend-builder

Builds UI components that are usable, accessible, and on-brand, covering loading, empty, and error states. Use when implementing user-facing interface work.

emtcmca/promptsmith · 34 tokens

designer

Designer agent for UI/UX patterns, design systems, and visual implementation. Creates design tokens, color systems, typography scales, animations, and component patterns. Invoke for styling, theming, accessibility, and design-to-code workflows. Works with Tailwind CSS 4, shadcn/ui, Framer Motion, and Figma.

shahtuyakov/claude-setup · 69 tokens