Profundis MCP server

The Profundis MCP server lets you query Profundis straight from an AI assistant. Instead of switching to the web UI or writing API calls by hand, you ask Claude, ChatGPT, or any Model Context Protocol client, and it runs the search and returns the results in your chat.

It exposes the same data as the web UI and the API: hosts, DNS records, WHOIS, TLS certificates, subdomain enumeration, ASN details, and perceptual favicon-similarity search.

What you can do with it

• Ask "what subdomains does example.com have?" and get the enumeration without leaving your assistant.
• Pivot from a host to its favicon, then find every site reusing a lookalike icon (the Brand Clone Radar flow).
• Build recon into an agentic workflow: your assistant chains search_hosts, search_dns, and domain_details to map a target.

How it works

The server is a thin proxy in front of the Profundis API. Your client connects over the MCP Streamable HTTP transport to:

https://mcp.profundis.io

Every request carries your Profundis credential. The server forwards it to the API, which authenticates the call, applies your plan's quotas, and bills credits exactly as a direct API call would. The MCP server holds no credential of its own.

Two ways to authenticate:

• OAuth (recommended): one-click "connect and sign in" from Claude, ChatGPT, and Claude Code. No key to copy or store.
• API key: paste your Profundis API key as a bearer token. Handy for scripts and headless clients.

See Authentication for the details, and Connect a client for the per-client steps.

What it costs

Usage bills your Profundis credits at the same rate as the equivalent API search: most searches cost 1 credit, a WHOIS search costs 3, and a favicon search costs 5 (see Credits). A paid plan or a Free account with purchased credits is required, the same as for the API. The full per-tool cost is listed in the Tools reference.