About
Enter a URL and A11y Check samples representative pages, audits them against WCAG 2.2 and KWCAG 2.2, and produces a report that shows what to fix and how. Automation is only the start: an evaluator adds manual verdicts on top, so the result is fit for real conformance reviews. It is free, and all of the code is public.
The story behind our name
알리첵 (Allycheck)
A11y Check
"a11y" is the standard numeronym for accessibility, counting the 11 letters between the a and the y. Read aloud in Korean it becomes "alli" (알리), which happens to sound exactly like the Korean verb for "to make known". Add "chek" (check) and you get our Korean name, 알리첵.
- 알리 (alli)
- The Korean reading of a11y — and the verb 알리다, "to make known". We surface issues that would otherwise stay invisible.
- 첵 (chek)
- Check: audit against the standards and tick issues off one by one, like the checkmark in our logo.
알리첵 — we make accessibility issues known, and check them off.
Why we built it
There are plenty of accessibility checkers. Most cover only the international standard or only the Korean one, and even the good ones leave you writing the actual report from scratch. We wanted a tool that evaluates both standards at once, produces a submission-ready report, and lets anyone inspect the judgment logic in code. So we built it.
Who's behind it
This service is built and run by the web manager at Daegu Cyber University.
Daegu Cyber University was founded on the spirit of "Love, Light, and Freedom" and has specialized in special education, rehabilitation, and welfare. When you run the web for a university whose mission is widening the door of learning for students with disabilities, accessibility stops being a compliance item and becomes the obvious thing to do. Allycheck is a tool that makes that obvious thing easy for other organizations and developers too.
About Daegu Cyber University ↗Who it's for
Web managers at organizations
See where your site stands in numbers, and hand the report to management or your vendor as a share link or PDF. Turn on scheduled scans and it watches your site daily.
Developers and agencies
Every violation comes with the page URL, CSS selector, offending code, and a fix example. Export an AI fix request file and paste it into Claude or ChatGPT to get corrected code.
Accessibility evaluators
Automation lays the groundwork; anything it can't judge is left as a manual check with testing instructions. Record verdicts and notes to complete a combined automated + manual report.
Standards and methodology
We didn't invent our own criteria. We follow what W3C and the Korean national standard define.
WCAG-EM 2.0
W3C's official accessibility conformance evaluation methodology (2.0, published July 2026). Scoping, exploration, representative sampling, auditing, and reporting follow this procedure, with both a W3C Report Tool-compatible export and a 2.0-referenced export.
WCAG 2.2
The international guidelines. Each of the 55 A/AA success criteria is shown as passed, violated, or needs manual review.
KWCAG 2.2
The Korean guidelines (KS X OT0003), presented as the same 33-item matrix used in domestic certification audits.
What is KWCAG 2.2?
KWCAG 2.2 — the Korean Web Content Accessibility Guidelines — is South Korea's national standard, organised as 33 checkpoints across four principles: perceivable, operable, understandable, and robust. Korea's accessibility quality certification is assessed against it. It overlaps heavily with WCAG 2.2, but the grouping, wording, and a few requirements differ, so a product built to WCAG alone can still fall short of a Korean review. If you ship a service used in Korea, checking both at once is the cheapest way to find that out early rather than late.
What it does
Automated audits
Samples pages from your sitemap and internal links, then runs 106 axe-core rules: alt text, color contrast, keyboard access — the things a machine can judge with confidence. Pages are re-checked at mobile width, and same-origin iframes are audited too.
Evaluation workbench
Evaluators record per-criterion verdicts and observations on top of the automated results. Add site details and an executive summary and it becomes the print/PDF report.
Chrome extension
Admin screens behind login and bot-blocked sites are out of a server's reach, so the extension audits them right in your browser and merges results into your site report. It also ships on-page tools: color-blindness simulation, focus-order overlay, a contrast eyedropper.
AI coding tool integration (MCP)
Register the MCP server (@a11ychk/mcp) in Claude Code or Cursor and the agent can scan your localhost pages in place, read per-violation remediation guides, and fix the code — using the same engine and judgment as this service.
Accessibility badge
Verified domains get an SVG badge showing the latest compliance rate, ready to embed.
Scheduled scans and alerts
Turn it on and your site is scanned daily. If compliance drops or new violations appear, you get an email; the dashboard shows the trend over time.
Honest automation
Automated tools can only find a portion of accessibility issues. We state that limit in the report itself, not in fine print, and mark everything requiring human judgment as a manual check with testing instructions. Our automated results alone cannot support a conformance claim — and you should be skeptical of any tool that says otherwise.
GitHub
The code and development history of the scan engine, the WCAG/KWCAG mapping, and the Korean remediation guides are on GitHub. If a verdict looks wrong, check the reasoning there — or propose an improvement.
github.com/IsaacEryn/a11ychk ↗