<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Syntropy Digital</title>
    <description>The latest articles on DEV Community by Syntropy Digital (@syntropydigital).</description>
    <link>https://dev.to/syntropydigital</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4133057%2Fec0fd325-5555-41b4-b005-d45d6afe0562.png</url>
      <title>DEV Community: Syntropy Digital</title>
      <link>https://dev.to/syntropydigital</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/syntropydigital"/>
    <language>en</language>
    <item>
      <title>A zero-config accessibility gate for GitHub Actions (axe-core, WCAG 2.2, one step)</title>
      <dc:creator>Syntropy Digital</dc:creator>
      <pubDate>Sat, 19 Sep 2026 15:19:51 +0000</pubDate>
      <link>https://dev.to/syntropydigital/a-zero-config-accessibility-gate-for-github-actions-axe-core-wcag-22-one-step-29jf</link>
      <guid>https://dev.to/syntropydigital/a-zero-config-accessibility-gate-for-github-actions-axe-core-wcag-22-one-step-29jf</guid>
      <description>&lt;p&gt;Most teams find out their site fails accessibility checks when a customer, a lawyer, or a Lighthouse score tells them. The fix for the &lt;em&gt;finding-out&lt;/em&gt; part is cheap: run axe-core on every pull request and fail the build on serious violations. This action does exactly that, in one workflow step.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it does
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Runs &lt;code&gt;@axe-core/cli&lt;/code&gt; against a URL you give it (your preview server, staging, or a static build).&lt;/li&gt;
&lt;li&gt;Tags the run with the standard you choose: &lt;code&gt;wcag2a&lt;/code&gt;, &lt;code&gt;wcag2aa&lt;/code&gt;, &lt;code&gt;wcag21aa&lt;/code&gt;, or &lt;code&gt;wcag22aa&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;Turns every violation into a GitHub annotation — rule id, the CSS selector of the failing node, and axe's failure summary — and writes a Markdown table into the job summary.&lt;/li&gt;
&lt;li&gt;Fails the job when violations at or above your threshold appear (&lt;code&gt;fail-on: serious&lt;/code&gt; by default; &lt;code&gt;critical&lt;/code&gt;, &lt;code&gt;moderate&lt;/code&gt;, or &lt;code&gt;minor&lt;/code&gt; if you want it stricter).&lt;/li&gt;
&lt;li&gt;Fails the job if the scan itself could not run (unreachable URL, no browser). A scanner that silently passes is worse than a red build, so that case is an error, not a warning.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  The whole setup
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight yaml"&gt;&lt;code&gt;&lt;span class="na"&gt;name&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;Accessibility&lt;/span&gt;
&lt;span class="na"&gt;on&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="pi"&gt;[&lt;/span&gt;&lt;span class="nv"&gt;pull_request&lt;/span&gt;&lt;span class="pi"&gt;]&lt;/span&gt;
&lt;span class="na"&gt;jobs&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt;
  &lt;span class="na"&gt;a11y&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt;
    &lt;span class="na"&gt;runs-on&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;ubuntu-latest&lt;/span&gt;
    &lt;span class="na"&gt;steps&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt;
      &lt;span class="pi"&gt;-&lt;/span&gt; &lt;span class="na"&gt;uses&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;actions/checkout@v4&lt;/span&gt;
      &lt;span class="pi"&gt;-&lt;/span&gt; &lt;span class="na"&gt;run&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;npm ci &amp;amp;&amp;amp; npm run build &amp;amp;&amp;amp; (npm run start &amp;amp;) &amp;amp;&amp;amp; npx wait-on http://localhost:3000&lt;/span&gt;
      &lt;span class="pi"&gt;-&lt;/span&gt; &lt;span class="na"&gt;uses&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;syntropydigital/axe-core-action@v1&lt;/span&gt;
        &lt;span class="na"&gt;with&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt;
          &lt;span class="na"&gt;url&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;http://localhost:3000&lt;/span&gt;
          &lt;span class="na"&gt;standard&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;wcag22aa&lt;/span&gt;
          &lt;span class="na"&gt;fail-on&lt;/span&gt;&lt;span class="pi"&gt;:&lt;/span&gt; &lt;span class="s"&gt;serious&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;That's it. The job summary shows a table of rules with impact badges and links to the axe rule documentation; the annotations point at the exact selectors.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it does not do
&lt;/h2&gt;

&lt;p&gt;Automated checks catch roughly half to two-thirds of accessibility defects (Deque's own estimate for axe-core is 57% of issues by count). Keyboard traps, focus order, meaningful alt text, and reading order still need a person. Treat this action as the floor, not the audit — it stops regressions and catches the mechanical failures (missing labels, contrast, ARIA misuse, duplicate ids) before review.&lt;/p&gt;

&lt;h2&gt;
  
  
  Inputs
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;input&lt;/th&gt;
&lt;th&gt;default&lt;/th&gt;
&lt;th&gt;meaning&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;url&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;—&lt;/td&gt;
&lt;td&gt;page to scan (required)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;standard&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;&lt;code&gt;wcag21aa&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;axe tag set: &lt;code&gt;wcag2a&lt;/code&gt;, &lt;code&gt;wcag2aa&lt;/code&gt;, &lt;code&gt;wcag21aa&lt;/code&gt;, &lt;code&gt;wcag22aa&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;fail-on&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;&lt;code&gt;serious&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;minimum impact that fails the job&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;output-json&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;&lt;code&gt;axe-results.json&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;where the full axe report is saved (upload it as an artifact if you want history)&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Source, MIT licence, issues: &lt;a href="https://github.com/syntropydigital/axe-core-action" rel="noopener noreferrer"&gt;https://github.com/syntropydigital/axe-core-action&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;If you want the manual half of the audit — keyboard flows, screen-reader checks, and code-level fixes for the failures — that is what we do for a living; the README has the contact.&lt;/p&gt;

</description>
      <category>a11y</category>
      <category>github</category>
      <category>actions</category>
      <category>webdev</category>
    </item>
  </channel>
</rss>
