// VYDAI MCP server

Ask about your brand's AI visibility in Claude, Cursor and Codex

Connect your VYDAI workspace to your AI assistant, and it will answer questions about your brand using your check data: visibility, competitors, sources, trends. No spreadsheet exports, no switching tabs.

MCP

MCP (Model Context Protocol) is an open standard that lets AI assistants use data from other services. VYDAI gives this access read-only.

Works withClaudeCursorCodex

// Sample questions

Plain questions, answers from your data

The assistant picks the VYDAI data it needs on its own. Here is what people usually ask.

“How did our brand visibility change over the last month?”

Visibility percentage, change from the previous period, and a breakdown by ChatGPT, Gemini, Claude and Google AI Overviews.

“Which competitors does AI mention more often than us?”

Mention share for your brand and competitors, and who grew or dropped over the period.

“Which prompts stopped mentioning us?”

Prompts where the brand appeared at the start of the period and disappeared by the end, with the AI answers.

“Which sites does AI cite most in our niche?”

Top sources, new and lost domains, and how often each one was cited.

“What did the latest GEO audit of our site show?”

Overall score, blockers and what to fix first.

“How many visitors came from AI assistants?”

Traffic from ChatGPT, Perplexity and other AI sources from Google Analytics 4, if it is connected in your workspace.

There are also two ready-made scenarios: a monthly visibility report and a lost-mentions review. They show up in the assistant as ready commands.

// Available data

What the assistant sees

Only what is already in your workspace, and only in the projects you chose for the key.

Read-only

The assistant cannot change prompts, run checks, delete data or spend your balance.

  • Projects, prompts and AI answers
  • Brand visibility, trends and check status
  • Breakdown by prompt and AI system
  • Competitors and their mention share
  • Sources that AI systems cite
  • GEO audit results for your site
  • Traffic from AI assistants from Google Analytics 4, if connected
  • Account balance and plan (for a key with access to all projects)
  • VYDAI public topic ratings

// Security

You decide what the assistant sees

Chosen projects only

You pick the projects when you create a key. The assistant does not see the rest.

Read-only

The assistant reads data but cannot change, delete or run anything.

Revoke a key any time

One click in your workspace, and the key stops working from the next request.

Keys are stored as a hash

We show the full key once, when you create it. The database keeps only its hash, which cannot be turned back into the key.

// Setup

Three steps, a few minutes

01

Choose your assistant

In your workspace, open "AI assistants" and choose Claude Code, Claude Desktop, Cursor or Codex.

02

Create a key

Name the key and choose the projects. VYDAI shows a ready command or config snippet with your key.

03

Check the connection

Ask the assistant: "Show my VYDAI projects". If it replies with a list, you are all set.

This is what the Claude Code command looks like
claude mcp add --transport http vydai https://vydai.inweb.ua/mcp --header "Authorization: Bearer <YOUR_KEY>"

Your key is filled in automatically when you create it in the workspace.

Documentation →

// Assistants

Where it works

Claude Code

One command in the terminal.

Claude Desktop

A ready snippet for the config file.

Cursor

A ready snippet for mcp.json.

Codex

Two commands in the terminal.

Next

One-click connection in claude.ai and ChatGPT is planned. Then you will not need to copy a key.

// Try it

Connect VYDAI to your assistant

Creating a key and connecting takes a few minutes. You can revoke the key any time.

// FAQ

Frequently asked questions

What is MCP?

Model Context Protocol is an open standard that describes how an AI assistant requests data from external services. Claude, Cursor, Codex and other tools support it. VYDAI is one of those services.

Do I need to know how to code?

No. For Claude Code you run one command; for Claude Desktop and Cursor you paste a ready snippet into the config file. VYDAI prepares all of it in your workspace. Claude Desktop requires Node.js to be installed on your computer. It is free and takes a few minutes to install.

Can the assistant change anything in my workspace?

No. Access is read-only: the assistant does not edit prompts, run checks or spend your balance.

Does the assistant see all my projects?

Only the ones you chose for the key. You can create separate keys with different projects for different assistants or colleagues.

Does it use my balance?

No. The assistant reads data VYDAI has already collected. Checks keep running under your plan as before.

Who else sees the data the assistant receives?

Data from VYDAI becomes part of your conversation with the assistant, and the assistant's provider processes it under its own terms. So connect only the projects you are comfortable sharing with that assistant.

What if my key ends up in the wrong hands?

Revoke it in your workspace and create a new one. The old key stops working from the next request.

Documentation →