Method and limits
What an AxeGuard scan means
AxeGuard is useful automated signal, not a full accessibility audit. It helps developers find machine-checkable issues quickly, then decide where manual review should focus.
How the scan works
1
Validate the URL
AxeGuard accepts a public http/https page it can reach and blocks private, internal or unsafe targets.
2
Load one page in Playwright
The scanner opens a headless Chromium page at desktop size. It does not crawl a full site.
3
Run axe-core
The open-source axe-core engine checks machine-detectable accessibility rules and returns violations, impact and affected elements.
4
Show triage output
Results are grouped by axe rule, ranked by impact and linked to Deque rule docs for deeper review.
Covered
- One reachable public page per scan.
- Machine-checkable axe-core violations.
- Affected selectors, HTML snippets and axe failure notes.
- WCAG labels where axe maps a rule to a success criterion.
- Basic failed-state categories such as timeout, unreachable page, bot protection or private network block.
Not covered
- Authenticated pages, private dashboards or internal network targets.
- A full-site crawl, sitemap scan or mobile viewport audit.
- Human judgments such as meaningful alt text, caption quality or whether page flow makes sense.
- Legal advice, accessibility certification or an assessment of conformance with any standard or law.
- Paid monitoring, CI checks or client summaries. Those are still validation hypotheses.
Data used by a scan
AxeGuard stores the submitted URL, scan status, axe result summary and validation events. Rate limiting uses a salted hash of the requester IP rather than storing the raw IP. Email is optional and only used for monitoring early access.