AI Visibility Check

AI Visibility Check MCP server

Audit how AI assistants and search engines find, understand and recommend a website, from inside Claude, ChatGPT, Cursor, VS Code or any other MCP client.

AI Visibility Check is a hosted service. This repository holds no code: it describes the remote MCP server at

https://aivisibilitycheck.io/mcp

for the MCP Registry (server.json) and for the people connecting to it (this file). There is nothing to install.

• Transport: Streamable HTTP
• Sign-in: OAuth 2.1 with PKCE and dynamic client registration. Your client opens our sign-in page; a free account can be created there.
• Registry name: io.aivisibilitycheck/mcp

What it does

Ask your assistant to audit a website and it runs the audit on your account, then answers questions about the result: what is wrong, what to fix first, and whether AI assistants recommend the business.

• Quick audit: checks the homepage in about a minute. Free.
• Full audit: crawls the site, renders it in a real browser, checks outside sources and asks AI models about the brand. Takes about seven minutes and uses one full audit from your account's allowance, so the server tells the assistant to check with you before starting one. A site that had a full audit in the last day gets that report back instead of a new run, unless you ask for a new one.

A new account comes with three full audits and gets one more each month. Nothing is sold through the MCP server: its answers never contain a price or a payment link.

Connect

Claude

1. Settings → Connectors → Add custom connector.
2. Name: AI Visibility Check, Remote MCP server URL: https://aivisibilitycheck.io/mcp.
3. Add, then Connect, and sign in on the page that opens.

Step by step: aivisibilitycheck.io/claude-connector

ChatGPT

1. Apps → Add → Create, if your plan allows custom apps.
2. Name: AI Visibility Check, Server URL: https://aivisibilitycheck.io/mcp, Authentication: OAuth.
3. Confirm and sign in on the page that opens.

Step by step: aivisibilitycheck.io/chatgpt-app

Claude Code

claude mcp add --transport http aivisibilitycheck https://aivisibilitycheck.io/mcp

Then run /mcp in Claude Code and choose Authenticate.

Cursor

~/.cursor/mcp.json, or .cursor/mcp.json in a project:

{

"mcpServers": { "aivisibilitycheck": { "url": "https://aivisibilitycheck.io/mcp" } } }

VS Code

.vscode/mcp.json:

{

"servers": { "aivisibilitycheck": { "type": "http", "url": "https://aivisibilitycheck.io/mcp" } } }

Other clients

Any client that speaks Streamable HTTP and MCP's OAuth flow can connect to https://aivisibilitycheck.io/mcp. It discovers the sign-in server from https://aivisibilitycheck.io/.well-known/oauth-protected-resource/mcp.

What to ask

• "Run a quick AI visibility audit of example.com."
• "Run a full audit of example.com for the keywords plumber leeds and emergency plumber."
• "What are the critical issues?"
• "Show me the structured data section."
• "What should I fix first, and how much would it raise the score?"
• "Do AI assistants recommend this business? Which competitors do they name instead?"
• "How many full audits do I have left?"

Tools

What it can access

Once you allow the connection, your client can start audits and read your audits and their results, your email address and your remaining allowance. It cannot buy anything, change your plan or payment details, or delete anything.

Disconnect any time with Disconnect under *Connected apps* on your account page, or by removing the server from your client. Access ends within a minute.

What you ask your assistant, and the results sent back to it, are also processed by that assistant's provider under its own terms. Our privacy notice covers what we do with your data; using the service means agreeing to our terms.

Support

info@cyberspark.uk. Report a problem with this listing as a GitHub issue.

AI Visibility Check is made by Cyber Spark IT Services LTD (company number 14891489).