Use case · MCP clients
Keep private data out of MCP tool calls.
Claude Desktop, Cursor and VS Code connect to tools through MCP (Model Context Protocol, the way these apps use tools). The Vrymdor™ AI Node is an MCP server: it masks what those tools send and receive, and shows the app only the tools your policy allows.
The problem
A tool call carries arguments the model wrote, and the tool’s answer comes back into the chat.
- A tool call’s arguments can include an email address or a key the model read earlier in the conversation.
- A tool’s answer, such as a database row or a file, comes back into the chat and goes to the model provider.
- A tool the model should never use stays in its list unless something removes it.
With the AI Node
Desktop apps start a local command, so they connect through the node’s stdio bridge, which runs from the node image in Docker.
- Download the node from your dashboard and load it into Docker.
- Add the vrymdor server to the app’s MCP settings with your Node Host URL, Agent ID and Agent Secret: claude_desktop_config.json for Claude Desktop, .cursor/mcp.json for Cursor, .vscode/mcp.json for VS Code.
- Restart the app. The Vrymdor tools appear in its tools menu.
- Ask it to use vrymnox_sanitize on: mail [email protected]. The tool answers mail [REDACTED:EMAIL].
Use test data for this first check, never real personal data.
Try it on a tool call’s arguments
The AI Node hides email addresses, phone numbers, card numbers, US Social Security numbers, private keys, and the API keys and tokens for OpenAI, Anthropic, Stripe, GitHub, AWS, Google and Slack, plus bearer tokens and JWTs. Each one becomes a label like [REDACTED:EMAIL], so the model still knows something was there.
The tool still gets a useful issue, without the address, the token or the card. Masked here: EMAIL 1, GITHUB_TOKEN 1, CREDIT_CARD 1.
This demo uses the Playground's Local engine: public, well known detectors set to the types the AI Node masks by default, in the AI Node's [REDACTED:TYPE] format. In your workflows, the Vrymdor™ AI Node does the masking.
What it does not cover
Plainly, so you know where you still need care.
- Names and street addresses are not masked by default. On Pro and Business you can add your own masking patterns for details with a fixed shape.
- It works by pattern matching on known formats, with the Luhn test for card numbers. A detail written in an unusual way, such as a card number spelled out in words, can get through.
- It hides private details before the model reads the text. It does not check what the model answers, and it does not make an AI model safe in general.
- Text you type into Claude Desktop or an IDE chat goes to that app’s model provider before any tool runs, so the AI Node cannot mask your own prompt. It protects what the tools send and receive.
Start free, upgrade when you grow.
Community is free for up to 2 agents and 10,000 requests a month, with a hard cap so nothing is ever billed. Pro is $29 a month for up to 15 agents and 100,000 requests, and Business is $99 a month for up to 50 agents and 500,000 requests, each with 14 days free. Enterprise is sized to your contract.