preferly

For agents: connect to https://www.preferly.ai/mcp (streamable HTTP, no auth). Server card: https://www.preferly.ai/.well-known/mcp/server-card.json · This page as Markdown: /ai-visibility-mcp-server.md

Preferly MCP server

Free AI visibility checks inside Claude, ChatGPT, Cursor and any other AI tool that speaks MCP. Ask your assistant whether AI crawlers can reach a website, whether its llms.txt and business schema are right, and whether AI can read its hours and prices, and it runs Preferly’s checks for you.

Server URL
https://www.preferly.ai/mcp

Streamable HTTP · No sign-up · No API key · Read-only

Set up in your AI toolTry the free check in your browser

Quick setup

One command adds Preferly to every MCP client it finds on your computer:

Terminal
npx add-mcp https://www.preferly.ai/mcp

Prefer to do it by hand? Pick your tool below.

Set up in your AI tool

Claude Code

Run:

Terminal
claude mcp add --transport http preferly "https://www.preferly.ai/mcp?client=claude"

Then ask Claude a question from Try asking below.

Claude (claude.ai and Claude Desktop)

Open Settings → Connectors → Add custom connector. Name: Preferly. URL: https://www.preferly.ai/mcp?client=claude. Leave authentication empty, save, then turn Preferly on from the tools menu in a chat. Custom connectors are available on Claude’s paid plans.

Connector URL
https://www.preferly.ai/mcp?client=claude

ChatGPT

Our ChatGPT app is in review. Until it’s listed, you can add Preferly yourself: in ChatGPT on the web, open Settings → Apps → Advanced settings, turn on Developer mode, then Create app. Name: Preferly. MCP server URL: https://www.preferly.ai/mcp?client=chatgpt-dev. Authentication: No authentication. Developer mode availability depends on your plan and workspace settings.

MCP server URL
https://www.preferly.ai/mcp?client=chatgpt-dev

Cursor

Add to Cursor

Or add this to ~/.cursor/mcp.json:

~/.cursor/mcp.json
{
  "mcpServers": {
    "preferly": {
      "url": "https://www.preferly.ai/mcp?client=cursor"
    }
  }
}

VS Code (GitHub Copilot)

Add to VS Code

Using VS Code Insiders? Add to VS Code Insiders.

Or run MCP: Add Server → HTTP and paste https://www.preferly.ai/mcp?client=vscode, or add to .vscode/mcp.json:

.vscode/mcp.json
{
  "servers": {
    "preferly": {
      "type": "http",
      "url": "https://www.preferly.ai/mcp?client=vscode"
    }
  }
}

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json:

~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "preferly": {
      "serverUrl": "https://www.preferly.ai/mcp?client=windsurf"
    }
  }
}

Codex

Run:

Terminal
codex mcp add preferly --url "https://www.preferly.ai/mcp?client=codex"

Gemini CLI

Add to ~/.gemini/settings.json:

~/.gemini/settings.json
{
  "mcpServers": {
    "preferly": {
      "httpUrl": "https://www.preferly.ai/mcp?client=gemini"
    }
  }
}

Goose

Run goose configure → Add Extension → Remote Extension (Streamable HTTP), name it preferly and use https://www.preferly.ai/mcp?client=goose.

Extension URL
https://www.preferly.ai/mcp?client=goose

Hermes Agent

Install the Preferly plugin:

Terminal
hermes plugins install preferly-ai/preferly-hermes --no-enable
hermes plugins enable preferly

Any other MCP client

Point it at https://www.preferly.ai/mcp (streamable HTTP, no auth). Add ?client= with your tool’s name so we can tell clients apart. If your client only runs local servers, use the bridge:

Terminal
npx mcp-remote https://www.preferly.ai/mcp

Tools

Every tool is read-only. Each works on one public website per call and changes nothing on it.

The Preferly MCP server’s tools
ToolWhat it doesReadsOn the web
check_ai_crawler_accessWhether a website's robots.txt lets AI crawlers in, whether its homepage loads for them or shows a bot check, whether it has /llms.txt, and its structured data types.The site's public pagesAI crawler check
generate_ai_robots_txtrobots.txt lines that let AI search and assistant crawlers in and keep training crawlers out, or any mix. Fetches nothing.NothingAI robots.txt generator
check_llms_txtWhether a website publishes /llms.txt, its size and title, and what doesn't follow the llmstxt.org format.The site's public pagesllms.txt checker
check_business_schemaThe business described in a homepage's structured data (JSON-LD): its type, which fields are present, missing or look off.The site's public pagesBusiness schema checker
check_business_infoWhether a website's name, phone, email, address, hours and prices are readable as text without running scripts.The site's public pagesCan AI read your business info?
quick_ai_visibility_checkAsks 3 AI models 3 typical customer questions about a business, from its public website, and reports whether each named it and where one contradicts the website.The site's public pagesFree check

The website checks fetch only public pages (robots.txt, the homepage, /llms.txt and at most 2 contact, booking or pricing pages), identifying themselves as PreferlyBot.

Try asking

  • “Can AI crawlers read example.com?”
  • “Does my robots.txt block GPTBot or ClaudeBot?”
  • “Write robots.txt lines that let AI search crawlers in, keep training crawlers out, and keep /admin private.”
  • “Does example.com have a valid llms.txt?”
  • “Is the business schema on example.com complete?”
  • “Can AI read the phone number, hours and prices on example.com without running scripts?”
  • “What do AI assistants say about the business at example.com?”

Limits, privacy and safety

Free, no sign-up and no key. Each caller gets 10 calls an hour and 30 website checks a day, and each website at most 20 fresh checks a day across all callers. The tools never log in, submit forms or change anything on a website. To apply these limits and count usage, we keep the website checked, a one-way hash of your IP address and the client name you add with ?client= for 7 days; see our privacy policy.

Free MCP checks or the full report?

The MCP tools check whether AI can reach and read a website. They don’t tell you what AI assistants say about your business. For that, Preferly’s full report asks 10 leading AI models about your business, quotes every place they contradict your website, and gives you the fixes ready to paste, for $99 once. See an example report.

Questions

Is the Preferly MCP server free?

Yes. There’s no sign-up, no API key and no charge. Free, no sign-up and no key. Each caller gets 10 calls an hour and 30 website checks a day, and each website at most 20 fresh checks a day across all callers.

Which AI tools can use it?

Any client that supports remote MCP servers over streamable HTTP, including Claude, Claude Code, ChatGPT developer mode, Cursor, VS Code, Windsurf, Codex, Gemini CLI, Goose and Hermes Agent. Clients that only run local servers can connect through mcp-remote.

Do I need to authenticate?

No. The server has no login and no OAuth. If your client asks for authentication, choose “None”.

What does it read on a website?

Only public pages: robots.txt, the homepage, /llms.txt and up to 2 contact, booking or pricing pages that robots.txt allows. It identifies itself as PreferlyBot and changes nothing.

Which AI crawlers does the crawler check cover?

10+ AI crawlers, including GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot and Google-Extended, and it shows the robots.txt rule that allows or blocks each one.

Why did a check return an older result?

A check for the same website in the last 10 minutes may be reused, so repeated questions are fast and don’t hit the site again.

Is it in the MCP Registry?

Yes, as ai.preferly/preferly in the official MCP Registry. Its server card is at /.well-known/mcp/server-card.json.

Is there a ChatGPT app?

Our ChatGPT app is in review. Until it’s listed, use developer mode as described above.

For developers and agents