Sentinel

Give your agenta fraud analyst.

Maskbreak speaks the Model Context Protocol. Point Claude, Cursor, or any MCP client at one URL and your agent can check any IP address for Tor exits and cloud-server hosting mid-conversation — no API key required for lookups. VPN and proxy verdicts, with the service named when known, come from a live visit through the REST API.

Connect it

One HTTP endpoint, no install, no package to pull. Add this to your MCP client configuration:

// Claude Desktop / Claude Code — claude_desktop_config.json
{
  "mcpServers": {
    "sentinel": {
      "type": "http",
      "url": "https://maskbreak.com/mcp"
    }
  }
}

Or from the Claude Code CLI:

claude mcp add --transport http sentinel https://maskbreak.com/mcp

Transport: JSON-RPC 2.0 over HTTP POST. Send Accept: application/json, text/event-stream. The server identifies as maskbreak v0.1.0 and advertises tools.listChanged.

The tools

Two, deliberately. Both are read-only and neither can mutate anything in your account.

lookup_ipKeyless

Screen an IP address for Tor exits, cloud-server hosting and anonymity. Returns an allow / review / block verdict, a 0–100 risk score, the individual signals, and network info (ASN, organization, country). The vpn and proxied fields are present but only fire on a live visit through the REST API.

service_statusKeyless

Current operational status of the Maskbreak API: uptime, per-service health, and measured processing latency. Useful for an agent that wants to check the source is healthy before trusting a verdict.

What an agent does with it

The interesting part is not the lookup, it's what the agent does around it. A support agent triaging a chargeback can check the order's IP without leaving the thread. A coding agent wiring up your signup flow can verify its own integration against a known Tor exit. An on-call agent investigating a traffic spike can classify the source in one step instead of pasting IPs into a web tool.

Try it in one line once connected:

You:  Is 185.220.101.34 a Tor exit node?
Agent: [calls sentinel.lookup_ip]
       Yes — Tor exit, risk score 90, verdict: block.

Beyond lookups

The MCP tools cover IP reputation. The full verdict pipeline — device fingerprinting, fake-browser detection, multi-account linking and the combined risk decision — runs over the REST API with a key, which stays free for 1,000 calls an hour.

Read the API docs →