Axe MCP Server One-click accessibility fixes
Connect Deque’s trusted accessibility expertise with AI coding agents to confidently build accessible code in your IDE.
What is the Axe MCP Server?
The Axe MCP Server is built on the Model Context Protocol (MCP) and brings Axe Platform expertise to AI coding agents directly in your IDE, allowing you to review, apply, or reject accessibility fixes without breaking flow.
Why use Axe MCP Server?
Accelerate fixes
Speed up remediation with accurate results and root-cause insights. Review and apply code fixes in one click.
Stay in flow
Get fix recommendations right in your IDE so you can code and correct without switching tools.
Grow expertise
Learn to build accessible code from the start with built-in testing, guidance, and learning.
Enterprise-grade security and data privacy
Get accessibility fixes instantly, right where you code.
Streamlined workflows
Get guidance and remediation fixes within your development environment to deliver accessible products, while eliminating context switching or long wait times for experts.
Context-specific guidance in your IDE
Work seamlessly in GitHub Copilot, Cursor, Claude Code, Windsurf, VS Code, and any AI coding assistant that supports MCP. The Axe MCP Server understands your unique workflow, delivering expert-backed, contextual code fixes.
Act on expertise you can trust
Get precise, accurate answers to even your most complex accessibility questions. Axe MCP Server is powered by the Axe DevTools engine and trained on Deque University’s comprehensive knowledge base.
“For years, we’ve talked about the promise of ‘shifting left’—finding and fixing accessibility issues earlier in the process. With technologies like this, real change happens.”
Jennison Asuncion, Co-Founder, Global Accessibility Awareness Day & the GAAD Foundation
Additional resources
Axe MCP Server: Digital accessibility expertise right in your AI agent
Supercharge your AI agents with built-in accessibility expertise for faster, accurate fixes.
Axe Assistant
Meet Axe Assistant, the gen AI chatbot trained on the industry’s most authoritative accessibility knowledge base.
2025 Forrester Digital Accessibility Platform Landscape and Deque Starter kit
Put insights into action with the 2025 Forrester Digital Accessibility Platform Landscape and Deque Starter kit.
Frequently asked questions
Is my data secure?
Yes, the Axe MCP Server is built with enterprise-grade protections and gives you full control of your data. Deque acts only as a data processor, never a controller, so your data is never used to train models. The result is trusted, expert-backed accessibility fixes with privacy and transparency at the core.
What platforms does the Axe MCP Server support?
The Axe MCP Server works with major IDEs and AI-powered coding assistants—including GitHub Copilot, Cursor, Claude Code, Windsurf, and VS Code—as well as any agent compatible with MCP. This lets you integrate accessibility testing directly into the tools you already use.
How does the Axe MCP Server fit in with the Axe DevTools Linter?
If you’re already using Axe DevTools Linter, you’re catching accessibility issues early by testing code in your IDE. The Axe MCP Server builds on this by introducing additional intelligence and context. It evaluates the rendered state of your application, providing actionable guidance in natural language based on specific prompts, all within the context of your IDE.
What sets the Axe MCP Server apart from other tools?
The Axe MCP Server uses the same trusted engine as the Axe DevTools Extension, ensuring rigorous accessibility testing while respecting your organization’s settings and policies.
Layered on top is Axe Assistant, Deque’s AI chatbot, which draws from Deque University, the most comprehensive digital accessibility knowledge base in the industry. Axe Assistant interprets user prompts and delivers consistent, expert-backed remediation guidance directly to the AI agent in your IDE.
This combination allows the Axe MCP Server to handle nuanced, specialized accessibility issues that more general tools might misinterpret, giving your teams more confidence in the guidance they receive.
