{
  "markdown": "# accessibility-scanner-mcp\n\nAn [MCP](https://modelcontextprotocol.io) server that lets an AI agent scan a web page for\n**WCAG accessibility issues** and get back findings it can act on. The agent calls one tool with a\nURL; it gets every violation grouped by severity, each with the **exact element selector, the\noffending HTML, the specific failure, the WCAG success criterion, and a fix-guide link** — plus the\nitems that still need human review.\n\nIt runs the real [axe-core](https://github.com/dequelabs/axe-core) engine in your **local Chrome**\n(via `playwright-core`), so nothing about the pages you scan leaves your machine. It scrolls the page\nfirst so **lazy-loaded content is actually checked** rather than silently skipped, and it resolves\ncolor contrast over CSS gradients, which most tools leave as \"needs review.\"\n\n> **Upgrading from 0.1.x?** 0.2.0 sees more of each page and returns more findings for the same URL.\n> See [CHANGELOG.md](CHANGELOG.md).\n\nPart of [accessibilityscanner.app](https://accessibilityscanner.app).\n\n## Requirements\n\n- Node.js 18+\n- Google Chrome installed (or set the `CHROME_PATH` environment variable to a Chromium binary)\n\n## Install\n\nAdd it to your MCP client's config. No global install needed — `npx` fetches it on first run.\n\n**Claude Desktop** (`claude_desktop_config.json`), **Cursor**, **Claude Code**, or any MCP client:\n\n```json\n{\n  \"mcpServers\": {\n    \"accessibility-scanner\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"accessibility-scanner-mcp\"]\n    }\n  }\n}\n```\n\nIf Chrome is not auto-detected, add an env block:\n\n```json\n{\n  \"mcpServers\": {\n    \"accessibility-scanner\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"accessibility-scanner-mcp\"],\n      \"env\": { \"CHROME_PATH\": \"/usr/bin/google-chrome\" }\n    }\n  }\n}\n```\n\n## The tool\n\n### `scan_accessibility`\n\n| Input | |\n|---|---|\n| `url` (string, required) | The http(s) URL to scan. |\n\nReturns a report grouped by severity. For each rule: the WCAG criterion, a fix-guide link, and per\nelement the selector, HTML, and exact failure. Example flow with an agent:\n\n> **You:** Audit https://example.com for accessibility and fix what you can.\n> **Agent:** *(calls `scan_accessibility`)* → reads the findings → edits the code → re-scans.\n\n## Honest about limits\n\nAutomated testing covers the machine-checkable subset of WCAG (most of the issues on a typical\npage, but not all of it). Items that need human judgement are returned under \"Needs manual review.\"\nIt never claims a page is \"compliant.\"\n\n## License\n\nMIT\n",
  "bytes": 2520,
  "sha": "9aeb8116d3a36eb21bae5805faae56de05888bfee7feae6e19f85cd81ad6e21b",
  "repo_slug": "bishop81/accessibility-scanner-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_bishop81_accessibility_scanner_d6465f39/readme"
}