How do I install the Wcag Auditor skill?

Run npx skillmds@latest add imyourboyroy/wcag-auditor in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Wcag Auditor safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Wcag Auditor?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Wcag Auditor free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Wcag Auditor?

ImYourBoyRoy (@imyourboyroy) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Wcag Auditor by imyourboyroy — Agent Skill

Name: Wcag Auditor
Author: ImYourBoyRoy

Wcag Auditor

Run the Portable Web Toolkit bundled WCAG accessibility evidence gate (Playwright + axe). Use for a11y audits on toolkit-managed Astro sites. Never resolve AI/wcag-auditor. Ensure Playwright Chromium is installed in the client project first. After every run, open and share the stakeholder dashboard HTML with the user.

ImYourBoyRoy Updated Sep 17, 2026

File contents

WCAG Auditor

Use only Web_Toolkit/wcag_auditor (never AI/wcag-auditor or sibling trees).

Evidence gate only — not a WCAG conformance certificate.

Prerequisites

Client project has a linked Web_Toolkit/ and *.site-profile.json.
Playwright peers in the client project (playwright, @axe-core/playwright).
Chromium browser binary (npx playwright install chromium). The CLI auto-installs these unless --skip-playwright-install.

Preferred commands

# With site profile (ephemeral config if no wcag-auditor.config.mjs)
node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs run \
  --site-profile ./<site>.site-profile.json \
  --base-url https://example.com

# Persist a checked-in config
node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs init --site-profile ./<site>.site-profile.json
node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs run --site-profile ./<site>.site-profile.json --base-url http://127.0.0.1:4321

Confirm: node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs core-path must end with Web_Toolkit/wcag_auditor.

After every run — share the dashboard

Locate wcag-audit-dashboard.html under the run output directory (CLI prints stakeholder dashboard → …).
Open it and share the path with the user (Cursor canvas is optional extra; the HTML dashboard is the durable artifact).
Prefer findings that show file:line — those are best-effort maps from selectors/classes into src/. Unmapped rows still list the CSS selector for search.
JSON/SARIF remain for tools; dashboard is for humans.

Rules

Pass --site-profile for toolkit sites.
Prefer live URL or local preview --base-url; do not invent routes.
Fix findings with smallest safe CSS/markup changes; re-run after fixes.
Do not treat PageSpeed Accessibility scores as WCAG proof.
Always surface the stakeholder dashboard path in the chat reply after a run.
Glassmorphism / frost UI is supported. Axe color-contrast incompletes map to cantTell (exit 3 unresolved), not an auto-fail. Do not strip frost UI solely to silence the gate. Prefer: raise frost opacity / solid underlay under text → verify AA with a contrast sampler → complete frost-glass-contrast in wcag-audit/manual-evidence.json → add a bounded suppression with outcomes: ["cantTell"] for axe/color-contrast on the route. See Web_Toolkit/wcag_auditor/docs/GLASSMORPHISM.md.
Never set gate.unresolvedEvidence: "ignore" just to ship glass; resolve evidence properly.

---name: wcag-auditordescription: Run the Portable Web Toolkit bundled WCAG accessibility evidence gate (Playwright + axe). Use for a11y audits on toolkit-managed Astro sites. Never resolve AI/wcag-auditor. Ensure Playwright Chromium is installed in the client project first. After every run, open and share the stakeholder dashboard HTML with the user.---# WCAG AuditorUse **only** `Web_Toolkit/wcag_auditor` (never `AI/wcag-auditor` or sibling trees).Evidence gate only — not a WCAG conformance certificate.## Prerequisites1. Client project has a linked `Web_Toolkit/` and `*.site-profile.json`.2. Playwright peers in the **client** project (`playwright`, `@axe-core/playwright`).3. Chromium browser binary (`npx playwright install chromium`). The CLI auto-installs these unless `--skip-playwright-install`.## Preferred commands```bash# With site profile (ephemeral config if no wcag-auditor.config.mjs)node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs run \ --site-profile ./<site>.site-profile.json \ --base-url https://example.com# Persist a checked-in confignode ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs init --site-profile ./<site>.site-profile.jsonnode ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs run --site-profile ./<site>.site-profile.json --base-url http://127.0.0.1:4321```Confirm: `node ./Web_Toolkit/wcag_auditor/bin/wcag-auditor.mjs core-path` must end with `Web_Toolkit/wcag_auditor`.## After every run — share the dashboard1. Locate `wcag-audit-dashboard.html` under the run output directory (CLI prints `stakeholder dashboard → …`).2. **Open it and share the path with the user** (Cursor canvas is optional extra; the HTML dashboard is the durable artifact).3. Prefer findings that show `file:line` — those are best-effort maps from selectors/classes into `src/`. Unmapped rows still list the CSS selector for search.4. JSON/SARIF remain for tools; dashboard is for humans.## Rules1. Pass `--site-profile` for toolkit sites.2. Prefer live URL or local preview `--base-url`; do not invent routes.3. Fix findings with smallest safe CSS/markup changes; re-run after fixes.4. Do not treat PageSpeed Accessibility scores as WCAG proof.5. Always surface the stakeholder dashboard path in the chat reply after a run.6. **Glassmorphism / frost UI is supported.** Axe `color-contrast` incompletes map to `cantTell` (exit 3 unresolved), not an auto-fail. Do **not** strip frost UI solely to silence the gate. Prefer: raise frost opacity / solid underlay under text → verify AA with a contrast sampler → complete `frost-glass-contrast` in `wcag-audit/manual-evidence.json` → add a bounded suppression with `outcomes: ["cantTell"]` for `axe/color-contrast` on the route. See `Web_Toolkit/wcag_auditor/docs/GLASSMORPHISM.md`.7. Never set `gate.unresolvedEvidence: "ignore"` just to ship glass; resolve evidence properly.

Frequently asked questions