WCAG Accessibility Scanner

Scan entire websites for WCAG 2.0, 2.1, and 2.2 accessibility issues with axe-core. Crawl multiple pages, find actionable accessibility problems, and export structured findings plus JSON and HTML summaries.

What it detects

The Actor runs axe-core in a real Chromium browser and can identify issues such as:

• Missing or invalid image alternative text
• ARIA roles, attributes, and accessible-name problems
• Color contrast failures
• Missing form labels
• Invalid heading and landmark structures covered by axe rules
• Keyboard/accessibility-related rule failures detectable by automation
• Link and button naming problems
• Table, language, document structure, and semantic HTML issues
• Additional axe best-practice checks in Comprehensive mode

Automated tools cannot prove WCAG conformance. Some accessibility requirements require manual testing with keyboard navigation, screen readers, zoom, reflow, cognitive review, and real user testing.

Why use this Actor

• Site-wide crawling instead of checking one page at a time
• WCAG 2.0 / 2.1 / 2.2 A + AA profiles
• Real browser rendering for JavaScript-heavy sites
• Confirmed violations + optional manual-review findings
• One result per rule per page to keep the dataset clean and avoid duplicate-heavy output
• Affected element samples with selectors, HTML snippets, and failure summaries
• JSON, CSV, Excel, XML, JSONL, and API exports through the Apify dataset
• HTML report + JSON summary for each run
• Crawl depth, page limit, URL include/exclude patterns, concurrency, and dynamic-page wait controls
• Protection against localhost/private-network scanning and common crawl traps

Input

A basic scan only needs a URL:

{
  "startUrls": [{ "url": "https://example.com" }]
}

A larger WCAG 2.2 AA scan:

{
  "startUrls": [{ "url": "https://example.com" }],
  "standard": "wcag22aa",
  "maxPages": 250,
  "maxDepth": 4,
  "sameDomainOnly": true,
  "includeSubdomains": false,
  "includeNeedsReview": true,
  "maxConcurrency": 5
}

WCAG profiles

Value	Checks
wcag2aa	WCAG 2.0 Level A + AA
wcag21aa	WCAG 2.0 + 2.1 Level A + AA
wcag22aa	WCAG 2.0 + 2.1 + 2.2 Level A + AA
comprehensive	WCAG 2.2 A + AA plus axe best-practice rules

Output

Each dataset row represents one accessibility rule on one page:

{
  "url": "https://example.com/contact",
  "pageTitle": "Contact",
  "issueType": "violation",
  "impact": "critical",
  "ruleId": "image-alt",
  "help": "Images must have alternative text",
  "description": "Ensure images have alternative text or a role of none or presentation",
  "helpUrl": "https://dequeuniversity.com/rules/axe/...",
  "wcagTags": ["wcag2a", "wcag111"],
  "allTags": ["cat.text-alternatives", "wcag2a", "wcag111"],
  "nodeCount": 3,
  "nodes": [
    {
      "target": ["img.hero"],
      "html": "<img class=\"hero\" src=\"hero.jpg\">",
      "failureSummary": "Fix any of the following..."
    }
  ],
  "scannedAt": "2026-09-18T12:00:00.000Z"
}

The run also stores:

• SUMMARY.json — totals, impact counts, unique rules, scanned pages, and failures
• REPORT.html — a simple human-readable run summary

Result types

violation means axe-core found an automated rule failure.

needs-review means axe-core could not determine the result automatically and a person should review it. Disable these with includeNeedsReview: false if you only want confirmed automated violations.

Crawl behavior

By default the Actor:

• stays on the starting hostname(s)
• scans up to 100 pages and 3 link levels
• removes query strings from discovered links to reduce duplicate/crawl-trap URLs
• skips common non-HTML file extensions
• removes URL fragments
• blocks localhost, link-local, private, reserved, and metadata network targets
• follows redirects while blocking redirects/subresources that attempt to reach private networks

Start URLs keep their query string, except common tracking parameters are removed.

URL filters

Use simple * wildcards:

{
  "includePatterns": ["*/blog/*", "*/products/*"],
  "excludePatterns": ["*/login*", "*/checkout/*"]
}

Filters apply to discovered links. Supplied start URLs are always scanned if they pass the public-URL safety checks.

Recommended use cases

• Accessibility audits before release
• WCAG regression checks after deployments
• Agency QA across client websites
• Finding recurring accessibility defects across large sites
• CI/CD or scheduled accessibility monitoring with Apify Tasks and Schedules
• Producing structured accessibility data for dashboards and remediation workflows

Technical stack

• Apify SDK
• Crawlee PlaywrightCrawler
• Chromium / Playwright
• Deque axe-core

The browser image and Playwright version are pinned for reproducible Apify builds.

Limitations

• Automated testing finds only issues that can be reliably evaluated in software; manual accessibility testing is still necessary.
• Cross-origin or sandboxed iframes may not be fully inspectable.
• Authenticated/private applications require a separate authenticated workflow; this Actor intentionally blocks private-network targets.
• Visual states that only appear after complex interactions may require purpose-built interaction steps.

Responsible scanning

Only scan websites you are authorized to test. Keep concurrency reasonable for the target website.