Search articles, authors and pages, or jump to a dashboard section
Governance
We target WCAG 2.2 Level AA. This page describes what's actually been verified and automated, not a claim of formal certification — there is no external body that certifies WCAG conformance the way COPE membership or Scopus indexing works.
Color contrast
Core text/background token pairs were computed directly (OKLCH → sRGB): 18.4:1 body text, 5.4:1–7.4:1 muted text, 10.7:1–7.6:1 primary — all pass AA's 4.5:1 minimum, in both light and dark theme.
Keyboard navigation
A skip-to-content link, a visible focus ring on every interactive element, and every click-to-navigate control (including card-style links) also responds to Enter/Space.
Labeled form fields
Every input on the submission form — including the repeating author and funder blocks — has a programmatically associated label, not just placeholder text.
Landmarks & document structure
Real <header>/<nav>/<main> landmarks and a declared page language, so assistive technology can navigate by region.
An axe-core accessibility scan runs in CI on every change, scoped for now to the pages that render without a database (most of this app has no distinct URL per page — it's a single-page app switching views under one address). It's currently informational rather than merge-blocking: axe-core has a documented limitation resolving contrast on gradient and glassmorphic backgrounds (it compares against the wrong composited color), which needs a human visual check to tell apart from a real issue — the same reason a rendered-page audit, not just a static one, matters.
AI-generated alt-text suggestions are produced automatically for every figure when an article publishes, but never applied automatically — an editor reviews and approves each one before it reaches the published galley.
There is no authoring tool that structures a manuscript's headings, alt-text, or semantic markup at submission time — submissions go through a plain text field, so accessibility of a published article's HTML depends on what the underlying conversion produces. Automated CI scanning only covers database-free pages today, not the full article-reading experience.
Tell us what you ran into and we'll look into it.