Introduction to Evinced MCP Servers

The Evinced MCP servers are Model Context Protocol (MCP) servers that enable AI agents and developers to integrate automated accessibility testing directly into their workflows. Built on Evinced's industry-leading accessibility scanning technology, they empower AI assistants to analyze web pages and mobile apps for WCAG compliance issues and provide actionable remediation guidance.

What are Evinced MCP servers?

Evinced MCP Servers act as a bridge between your AI coding agent and Evinced's accessibility engines. Once installed, your AI assistant gains the ability to:

• Detect accessibility violations — including keyboard navigation issues, semantic role problems, missing accessible names, color contrast failures, and more
• Prioritize issues — automatically sorted by severity (Critical, Serious, Moderate, Minor) so the most impactful problems are addressed first
• Remediate with guidance — receive context-aware, LLM-optimized instructions for fixing each detected issue type
• Validate fixes — re-scan after remediation to confirm issues are resolved and no regressions were introduced

Available MCP servers

Web MCP server

Integrates accessibility testing for web applications directly into your AI-driven development workflow.

• Analyzes web pages via URL using a headless Chromium browser
• Supports authenticated pages and complex states via a companion browser extension
• Works with any public or internal web application
• Compatible with Cursor IDE, VS Code, Claude Code, and any MCP-compatible client

View Web MCP Server documentation →

Mobile MCP server

Integrates accessibility testing for iOS and Android applications directly into your AI-driven development workflow.

• Analyzes live app screens at runtime on simulators, emulators, or physical devices
• Supports iOS (via WebDriverAgent) and Android (via ADB)
• Works alongside mobile automation MCPs like XcodeBuildMCP and Mobile MCP
• Compatible with Cursor IDE, Claude Code, and any MCP-compatible client

View Mobile MCP Server documentation →

Common use cases

1. Development Workflow Integration — Enable AI coding assistants to proactively identify and fix accessibility issues during feature development
2. Remediation Orchestration — Leverage AI agents to systematically analyze, prioritize, and remediate accessibility issues across entire applications
3. CI/CD Integration — Automate accessibility checks as part of your build and deployment pipeline

Getting started

Choose the MCP Server that matches your platform:

Both servers require an Evinced account and credentials. Contact your Evinced representative to obtain your service account ID and API key.

Additional resources

• Evinced Knowledge Base
• Model Context Protocol specification