# Accessibility MCP server for ChatGPT, Claude, Gemini and Cursor
URL: https://pandus.online/en/accessibility-mcp-server
Updated: 2026-10-20

Connect ChatGPT, Claude, Gemini, Cursor or VS Code to Pandus over MCP: explain any WCAG rule for free, see what to fix on your site and check the fix again.

AI assistants already write a large part of the code of many sites. The Pandus MCP server gives them what they lack for accessibility: **what is actually wrong on your pages**, **why it matters for your customers** and **how to fix it on your platform**. The assistant reads the problem, finds the template that renders it in your project, fixes the cause and starts a new check to confirm it.

It works with **any assistant that supports remote MCP servers**, not with one vendor. Results come from **an automated check, not a full audit**, and **not legal advice**.

## What the assistant can do

| Tool                     | What it does                                                                                                     | Needs            |
| ------------------------ | ---------------------------------------------------------------------------------------------------------------- | ---------------- |
| `search_rules`           | Finds rule cards by words, an axe-core rule id or a WCAG criterion                                               | nothing          |
| `explain_rule`           | One rule: what is wrong, who is affected, how to fix it in code and on Shopify, WooCommerce, PrestaShop, Shoptet | nothing          |
| `explain_wcag_criterion` | A WCAG 2.2 level A or AA criterion in plain words, with manual tests                                             | nothing          |
| `list_sites`             | Your sites with the latest check and problems by severity                                                        | API token        |
| `what_to_fix`            | Open problems of a site, most important first, from the scanner, real visitors, the extension and reports       | API token        |
| `get_problem`            | One problem: pages, CSS selectors, HTML examples, the fix for your platform, WCAG clauses                        | API token        |
| `get_report`             | The latest check: problem types, pages, trend                                                                    | API token        |
| `compare_checks`         | What was fixed, what is new and what is still open between the two latest checks                                | API token        |
| `start_check`            | Starts a new check of the site within your plan                                                                  | token with write |
| `check_status`           | The progress of a check                                                                                          | API token        |

Two ready prompts come with it: **fix the most important problems** of a site in your codebase, and **write an update for the site owner** in their language.

## Connect your AI assistant

The server address is `https://pandus.online/mcp` (streamable HTTP). Without a token it serves the free rule reference. For your own sites, create an API token in Pandus under **API and webhooks**: a read-only token is enough to read; add "write" to let the assistant start checks. Then add the server to your assistant.

### Claude (web and desktop)

In Claude, open **Customize → Connectors**, choose **Add custom connector**, enter the address and, for your own sites, add the request header `Authorization: Bearer YOUR_TOKEN`. On Team and Enterprise plans an owner adds the connector in the organization settings.

### Claude Code

```bash
claude mcp add --transport http pandus https://pandus.online/mcp --header "Authorization: Bearer YOUR_TOKEN"
```

### ChatGPT

On paid plans, turn on developer mode in the settings for apps and connectors, then create a connector with the address `https://pandus.online/mcp`. Without authentication you get the free rule reference. To read your own sites, use token (bearer) authentication with your API token where your plan offers it. On Business and Enterprise workspaces an admin may need to allow custom connectors first.

### Gemini CLI

```bash
gemini mcp add --transport http pandus https://pandus.online/mcp --header "Authorization: Bearer YOUR_TOKEN"
```

Or in `~/.gemini/settings.json`:

```json
{
    "mcpServers": {
        "pandus": {
            "httpUrl": "https://pandus.online/mcp",
            "headers": { "Authorization": "Bearer YOUR_TOKEN" }
        }
    }
}
```

### Cursor

In `~/.cursor/mcp.json` or `.cursor/mcp.json` in the project:

```json
{
    "mcpServers": {
        "pandus": {
            "url": "https://pandus.online/mcp",
            "headers": { "Authorization": "Bearer YOUR_TOKEN" }
        }
    }
}
```

### VS Code

In `.vscode/mcp.json`:

```json
{
    "servers": {
        "pandus": {
            "type": "http",
            "url": "https://pandus.online/mcp",
            "headers": { "Authorization": "Bearer YOUR_TOKEN" }
        }
    }
}
```

### Perplexity and other assistants

Perplexity (Pro, Max and Enterprise) adds remote MCP servers under **Settings → Connectors** as a custom remote connector. Windsurf, Cline and other MCP clients take the same address; where they ask for authentication, send your token as `Authorization: Bearer YOUR_TOKEN`.

Keep tokens out of shared repositories: use your assistant's user settings or an environment variable.

## What to ask

- "Use Pandus: what are the three most important accessibility problems on shop.example, and where are they in this repository?"
- "Fix the image-alt problem Pandus found on the product pages, then start a new check."
- "Explain WCAG 2.4.7 and check our button styles for it."
- "Compare the last two Pandus checks and write a short update for the client in German."

## Pandus and open-source axe MCP servers

| What you need                         | Pandus MCP server                                         | Open-source axe MCP servers                 |
| ------------------------------------- | --------------------------------------------------------- | ------------------------------------------- |
| Check one page now                    | Weekly checks of the whole site, plus a check on demand   | Yes, from your computer                     |
| What to fix first                     | Ranked by impact on customers, pages and real visitors    | A list of axe results                       |
| Barriers real visitors met            | Yes, with the Pandus SDK, as totals                       | No                                          |
| Fix for Shopify, WooCommerce and code | Yes, in 7 languages                                       | axe-core help texts                         |
| Before and after                      | Comparison of two checks                                  | You compare by hand                         |
| Pages behind a login, localhost       | With the [browser extension](https://pandus.online/en/accessibility-checker-browser-extension)       | Yes                                         |

Both kinds can run side by side: an open-source server for local tests while you develop, Pandus for the whole site and the team.

## Security and data

- **Only your team's data.** A token sees the sites of its own team, as in the [API](https://pandus.online/en/accessibility-testing-api), and stops working when the plan no longer includes the API.
- **Untrusted HTML.** Code snippets from your pages are shortened and marked as untrusted, so text hidden in a page cannot give the assistant orders.
- **No visitor data.** Barriers real visitors met come as totals; session recordings are never sent.
- **Limits and history.** Requests are limited per token, and every call over your data is written to your team's activity log.

## Plans

The rule reference is free for everyone. Your sites, "what to fix", reports and new checks over MCP come with the **Pro** and **Agency** plans, the same plans that include the API. See [pricing](https://pandus.online/en/pricing).

Q: What is an accessibility MCP server?
A: MCP (Model Context Protocol) is an open standard that lets AI assistants call the tools of other services. The Pandus MCP server gives an assistant accessibility rule cards, WCAG 2.2 explanations and, with your key, the check results of your own sites.

Q: Which AI assistants can use the Pandus MCP server?
A: Any client that connects to a remote MCP server over HTTPS (streamable HTTP): Claude on the web and in Claude Desktop, Claude Code, Cursor, VS Code, Gemini CLI, and ChatGPT and Perplexity on their paid plans. To read your own sites the client has to send your API token in the Authorization header.

Q: Does it work with ChatGPT?
A: Yes for the free rule reference: add https://pandus.online/mcp as a custom connector without authentication. Reading your own sites needs your API token as a bearer token; use it where your ChatGPT plan offers token authentication for custom connectors.

Q: Does it work with Gemini?
A: Yes, in Gemini CLI: add the server with the http transport and your token in the Authorization header.

Q: Is the Pandus MCP server free?
A: The rule reference is free and needs no account: search rules, explain an axe-core rule with fixes for code and shop platforms, and explain WCAG 2.2 criteria. Your sites, what to fix, reports and new checks need an API token of a Pro or Agency plan.

Q: How is it different from open-source axe MCP servers?
A: Open-source servers run axe-core on one page from your computer. Pandus adds the weekly check of the whole site, problems ranked by impact on customers, barriers real visitors met, fixes for your platform in seven languages and a comparison with the previous check.

Q: Can the assistant change my site through Pandus?
A: No. Pandus only reads results and, with a "write" token, starts a check. The assistant changes the code in your own project, where you review the changes before you deploy them.

Q: What about text hidden in my pages that tries to give the assistant orders?
A: HTML from checked pages reaches the assistant shortened, in fields marked as untrusted, and the server tells the assistant to treat it as data, never as instructions.

Q: Is a fix made with an AI assistant enough for WCAG or the EAA?
A: No. Pandus runs an automated check, which finds part of the problems, and is not legal advice. Test the changed parts with a keyboard and a screen reader as well.