skills.rest
Jin9
Jin9COMMUNITY
Explore all skills by this author →
GitHub

running-accessibility-tests

Run axe-core accessibility scans against built front ends and emit WCAG 2.1 AA gate verdicts.

Updated Apr 26, 2026
Search Tags:#accessibility#wcag#playwright#axe-core#a11y testing#gate verdict#front-end validation
One-click install
npx skills add https://github.com/Jin9/skillify-foundation --skill running-accessibility-tests-jin9
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: running-accessibility-tests
Source: https://github.com/Jin9/skillify-foundation/tree/main/treasury/running-accessibility-tests
Command: npx skills add https://github.com/Jin9/skillify-foundation --skill running-accessibility-tests-jin9

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires axe-core, playwright, and includes references (resource) components.

What problem does it solve? Manual accessibility sign-off is slow and automated tools alone are misleading, since engines like axe-core catch only about 57% of WCAG issues. This Skill runs automated WCAG 2.1 AA scans on a built front end and produces an honest PASS, FAIL, or ERROR gate that also surfaces the items requiring human review. ## Core Features & Use Cases - Automated WCAG 2.1 AA Scanning: Runs axe-core via a Playwright browser driver against pages, components, or a served build URL in a sandbox. - Honest Gate Verdicts: Emits PASS, FAIL, or ERROR with structured violations, needs-review items, and required manual checks, never claiming full coverage. - Banking-Grade Guardrails: Idempotent, report-only operation with PII redaction, audit IDs, and escalation to human queues on failure. - Use Case: During front-end review or SIT, point the Skill at your built pages to get an accessibility gate with machine-decidable violations plus a checklist of manual checks (keyboard traversal, focus order, alt text quality) for a named human to sign off. ## Quick Start Run accessibility tests on my built front end at the staging URL and produce the WCAG 2.1 AA gate verdict with violations and manual checks.

Frequently Asked Questions about running-accessibility-tests

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run automated WCAG accessibility tests on my front end?▼

Provide the built front end as pages, components, or a served build URL, and the Skill runs axe-core through a Playwright browser driver in a sandbox. It returns a PASS, FAIL, or ERROR verdict with violations, needs-review items, and manual checks.

What percentage of WCAG issues does axe-core catch?▼

Automated engines like axe-core catch about 57 percent of WCAG issues on average. The remaining criteria, such as meaningful alt text and logical focus order, are flagged as needs-review items requiring human verification rather than silently passed.

Can axe-core results be treated as full WCAG AA conformance?▼

No. A clean automated run is only the floor, not full conformance. The Skill maps engine incomplete results to needs_review and lists manual_checks like keyboard-only traversal, and unresolved items produce a FAIL under the banking gate policy.

Does this tool fix accessibility issues it finds?▼

No. The Skill is report-only and never edits the UI, tokens, or remediation code. Violations are routed to a front-end fixer, and manual checks are resolved by a named human who signs off on the gate.

What happens when the build cannot render during an accessibility scan?▼

If the build cannot render or the engine cannot run, the Skill emits an ERROR verdict and escalates to the human queue. If there is no reachable built front end at all, it stops with a needs-input status instead of producing a verdict.

Related Skills

Explore More Software Engineering Skills→
Adelie-SquadAdelie-Squad
17

Creative Frontend Agent

Transform design specifications into responsive React, Next.js, or Vue front-end implementations.

Official
Advanced
DaBigHomieDaBigHomie
1

human-approval-gate

Require human verification or independent reviewer approval for critical agent decisions.

Community
Advanced
ConductionNLConductionNL
1

review-pr

Reviews GitHub pull requests with inline comments and formal approve or request-changes verdicts.

Official
Advanced
dinoudondinoudon

accessibility-audit

Audit web applications for WCAG 2.2 compliance across seven accessibility categories.

Community
Advanced