Accessibility Scanner MCP
Web Accessibility Automated Scanning CLI Tool
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: Accessibility Scanner MCP. It is used for: Web Accessibility Automated Scanning CLI Tool Full Skill content: https://321skill.com/skills/accessibility-scanner-mcp/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
This Skill addresses the inefficiency of manual web accessibility compliance checks. In practice, developers often need to manually verify if pages meet WCAG standards, a process that is time-consuming and prone to oversight. This tool combines Playwright's browser automation capabilities with the Axe-core accessibility rule engine to enable one-click, comprehensive scanning.
Usage is straightforward: simply run npx mcp-accessibility-scanner interactive in your terminal to launch the interactive REPL, then use natural language commands to perform operations. For example, enter browser_navigate {"url":"https://example.com"} to navigate to the target page, then input scan_page {"violationsTag":["wcag2aa"]} to trigger the scan, which will return a list of violations and repair suggestions. It also supports features like keyboard auditing and screenshot export.
It is well-suited for front-end development teams and test engineers who need to quickly pass WCAG 2.0/2.1 compliance acceptance, especially for frequently iterated web applications or projects in sectors with strict accessibility requirements, such as government or education. AI Agents can also directly integrate this tool to dynamically check page accessibility within automated pipelines.
It is recommended to schedule full-site scans regularly within CI/CD workflows, utilizing the --headless mode for headless execution. Note that it primarily relies on browsers supported by Playwright (Chrome/Firefox/WebKit) and performs well with dynamically rendered pages (e.g., SPAs), but may not cover all edge cases for every JavaScript framework.
Key Features
Compared to a pure Axe-core CLI, it integrates Playwright's browser automation capabilities, eliminating the need to manually write test scripts for navigation, screenshots, and user interaction simulation, and supports interactive invocation by AI Agents via the MCP protocol. Compared to Lighthouse's accessibility audits, it offers more granular control (e.g., custom viewport sizes, device emulation) and an interactive REPL mode.
Limitations
Requires Node.js 18+ environment. The runtime automatically downloads Playwright browser binaries (approximately 200-400MB), imposing certain demands on network bandwidth and disk space.
FAQ
How do I run it in headless mode?
Simply add the `--headless` flag, e.g., `npx mcp-accessibility-scanner --headless interactive`.
Which browsers are supported?
Supports Chrome, Firefox, WebKit, and MsEdge, specified via the `--browser` parameter.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/accessibility-scanner-mcp/raw/index.md to read the original Skill definition (Markdown format) for Accessibility Scanner MCP, and install it according to the instructions.
Raw Markdown URL for AI: /skills/accessibility-scanner-mcp/raw/index.md