MCP Server

Getting Started

We host it at https://mcp.internetdata.io/mcp, and you sign in to it with your InternetData account. It also runs on your own machine from npm. Either way it sees where your organization's license for each database stands. Databases are licensed by contract, except the Open ones, tor_ip_v1 among them, which any organization downloads under CC BY-SA 4.0. For the rest, see the API documentation, or request an evaluation if your organization doesn't have a license yet.

In Claude

In claude.ai, the desktop app or Cowork, add https://mcp.internetdata.io/mcp as a custom connector and choose Use Claude's published identity when asked. In Claude Code, install our plugin, which adds the server with skills:

/plugin marketplace add internetdata/claude-plugin
/plugin install internetdata@internetdata

Either way you sign in with your InternetData account, and Claude never sees your API key. The steps, the skills and how to disconnect: docs.internetdata.io/integrations/claude.

In ChatGPT and Codex

In ChatGPT on the web, add https://mcp.internetdata.io/mcp as a custom MCP server under Plugins and choose OAuth. In Codex, install our plugin, which adds the server with skills:

codex plugin marketplace add internetdata/openai-plugin
codex plugin add internetdata@internetdata
codex mcp login internetdata

Either way you sign in with your InternetData account, and neither sees your API key. The steps: docs.internetdata.io/integrations/chatgpt and docs.internetdata.io/integrations/codex. On the OpenAI API, the Responses API's mcp tool takes the server as its server_url and your key as its authorization.

In any other MCP client

Point it at https://mcp.internetdata.io/mcp. A client that supports MCP authorization signs in the same way. One that doesn't can send a key instead, as Authorization: Bearer your-key.

On your own machine

You need an API key carrying the db.download scope. Add this to your MCP client's config:

{
  "mcpServers": {
    "internetdata": {
      "command": "npx",
      "args": ["-y", "internetdata-mcp"],
      "env": { "INTERNETDATA_API_KEY": "your-key" }
    }
  }
}

Requires Node.js 22 or newer. INTERNETDATA_BASE_URL overrides the endpoint if you need to point somewhere else.

Tools

Tool	What it answers
`list_databases`	Every database we publish, with its `standing` for your organization (`licensed`, `expired` or `unlicensed`) and the type and term of any license.
`database_metadata`	A database's columns, sample rows, row count, build date and file sizes.
`database_checksum`	The published digests for one database file.
`list_downloads`	Your organization's recent download attempts, refusals included.

Every tool is read-only.

Two spellings of a database id

list_databases answers a base id and a versions array:

{
  "base": "vpn_ip",
  "standing": "licensed",
  "open": false,
  "versions": [{ "id": "vpn_ip_v1", "version": 1, "formats": ["csvgz", "mmdb"] }]
}

The base is what a license names. versions[].id is what a download names, and it is the one database_metadata and database_checksum accept. The tools say so in their own descriptions, so an agent generally gets this right on its own; it is worth knowing when you read a transcript where one was refused.

There is no download tool, deliberately

The published builds run to several GB, which is not something an agent should pull into a conversation. database_metadata is how you find out what a transfer would cost, and the client libraries or the API are how you actually move the bytes.

Reading the download history

list_downloads is a bounded window - at most 200 attempts, newest first - and it lists refusals alongside successes, because a denial and its http_status are what answer "it stopped working". A database missing from the answer means it is not in that window, never that it was never fetched.

On this page