Log inSubmit server
Webability logo

Webability

Accessibility scanning, framework-aware fixes, visual audits, and regression checks for AI coding agents.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Webability

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "webability": {
      "command": "npx",
      "args": [
        "-y",
        "-p",
        "@webability/mcp",
        "webability-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Automated check passed— started and listed 20 tools correctly (Today).

Overview

webability MCP server scans live pages, HTML fragments, and multi-page journeys for WCAG accessibility issues using WebAbility detectors and axe-core. It returns structured findings with severity, selectors, fix operations, and optional source pointers, then supports framework detection, fix generation, verification, and baseline diffs. Use the hosted HTTP endpoint for public sites or the local stdio package for localhost and private development servers. Visual audits and full report deliverables require a free account on the hosted service.

Use cases

  • Scan a deployed page for WCAG violations
  • Check a localhost page during frontend development
  • Generate framework-aware accessibility fixes
  • Verify that a corrected element no longer fails
  • Compare a page against a previous accessibility scan

Key features

  • Dual-engine WCAG scanning with axe-core
  • Multi-page journey scans with deduplicated findings
  • Framework-aware fix generation
  • Fix verification and regression diffs
  • Contrast, ARIA, and visual audits
  • PDF, report, and Excel audit outputs

Tools20

What Webability lets your agent do. Expand a tool for its parameters.

Verified live~6.6k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 2d ago

Recent health check succeeded.

8/8 checks healthy over the last 55d

What it can do

From the 20 tools this server returned when we ran it; 20 declare their own safety hints.

Can delete or modify data · 1Handles credentials · 1Makes network requests · 15Read-only tools · 19
Network at startup3 connections
Files written at startupNone
Install scriptsNone
npm provenanceSigned build
Package age191 days
Maintainers1

Observed 2026-11-02 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 6d ago via OSV.dev · @webability/mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

npm downloads1.6kPackage downloads in the last 30 days.
Last commit3d agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed20Callable tools this server registers over MCP.

How Webability works

What webability does

webability brings accessibility checks into an MCP-compatible coding workflow. Its scanning tools inspect URLs, raw HTML, and multi-page journeys for WCAG issues, combining WebAbility detectors with axe-core and deduplicating overlapping findings. Results include issue summaries, selectors, severity, WCAG references, and structured fix operations such as adding, changing, or removing attributes and elements.

The server covers automated checks including contrast, labels, landmarks, ARIA, link consistency, target size, keyboard traps, and other WCAG 2.2 baseline rules. It can also inspect rendered output for visual concerns such as icon contrast, focus visibility, and controls that look interactive without behaving like buttons.

How it works

scan_page runs both scanning engines in parallel. Browser-based scans can expose source locations from React and Vue development builds, including the file, line, column, and component. scan_html handles an HTML snippet in process with jsdom; selecting its browser engine enables contrast checks that depend on rendering.

A typical workflow is find, fix, and verify. Use framework detection to identify Tailwind, MUI, Bootstrap, Next.js, WordPress, or plain CSS, then ask generate_ai_fix for alternatives appropriate to that framework. verify_fix rescans a selected element and reports whether the violation remains. diff_scan compares a baseline with the current page and separates fixed, new, and remaining findings.

The hosted endpoint is https://mcp.webability.io/mcp. It can scan public URLs without an API key. The local package runs through stdio, launches the browser on the developer’s machine, and can reach localhost and private addresses. Hosted scans cannot directly access those addresses; a tunnel is available for remote or hosted workflows.

Setup and configuration

For Claude Code, install the WebAbility plugin or add the hosted endpoint with an HTTP MCP transport. Cursor and VS Code provide install links, and other MCP clients can add the hosted URL as a custom connector. A local setup uses npx -y -p @webability/mcp webability-mcp as a stdio server.

Basic local scans do not require an account. Local AI judging is disabled until OPENROUTER_API_KEY is set; the hosted service performs that judging on its side. WEBABILITY_JUDGE_BASE_URL, model variables, timeout settings, and WEBABILITY_JUDGE=off control local judging. Optional PostHog variables enable or disable MCP usage analytics, and WEBABILITY_API_URL can point local operation at another backend.

Tools and capabilities

  • Scan URLs, HTML fragments, and multi-page journeys with configurable WCAG, rule, impact, and output filters.
  • Detect common frontend frameworks and generate fixes with mechanical, contextual, or visual fixability guidance.
  • Check color contrast and ARIA attributes independently.
  • List available rules and map selectors to local source files when framework source pointers are unavailable.
  • Keep local scan history and create a branded PDF from scan findings.
  • Run visual audits and start or poll full report and Excel workbook audits on the hosted service with a free account.

Limitations and notes

Automated results cannot determine whether alt text is meaningful or whether a custom widget is usable with a screen reader. Assistive technology testing is still needed. The server drops findings it cannot settle rather than presenting them as confirmed violations.

The hosted service applies fair-use limits of 30 browser scans per hour and 10 AI fixes per hour per IP. start_audit runs on the service and cannot reach a localhost URL, including when invoked through the local transport. Tunnel URLs require a separate secret, accept only GET and HEAD requests on the selected port, and expire after 30 minutes, five minutes of inactivity, or process termination.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Webability

webability is an MCP server that connects AI coding agents to accessibility scanning and remediation tools for web pages and HTML. Its main tools scan URLs, journeys, or snippets; detect frameworks; generate fixes; check contrast and ARIA; verify changes; and compare scans for regressions.

Install it by adding https://mcp.webability.io/mcp as an HTTP MCP server, or run npx -y -p @webability/mcp webability-mcp for a local stdio server. Claude Code, Cursor, and VS Code have dedicated installation options in the project setup.

Yes, the local stdio installation can scan localhost and private addresses because its browser runs on your machine. The hosted endpoint cannot reach your machine directly; use webability-tunnel or another tunnel when a hosted or remote workflow must scan a development server.

Yes, webability works with Cursor through its hosted MCP installation link or by adding https://mcp.webability.io/mcp as an HTTP server. Cursor can then call the scanning, fixing, verification, and other tools exposed through the endpoint.

No API key is needed for hosted scans and checks. Local AI judging requires OPENROUTER_API_KEY unless judging is disabled, while visual audits and full audit deliverables require a free hosted account.

Related MCP servers

All in Developer Tools All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewWebability AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/webability?style=directory)](https://allmcps.com/mcp/webability)
HTML Embed
<a href="https://allmcps.com/mcp/webability"><img src="https://allmcps.com/api/badge/webability?style=directory" alt="Webability on AllMCPs" /></a>