badchars/darknet-mcp-server ? reverse-engineered prompt

Build me an MCP server in TypeScript that gives an AI assistant access to dark web and threat intelligence from one place.

I want it to connect to common sources for breach data, ransomware tracking, Tor .onion access, malware and hash lookups, blockchain intel, exploit search, stealer logs, and other threat feeds, then expose all of that as easy to use tools the agent can call in conversation. It should work out of the box with free sources, and support optional API keys for premium lookups.

Please make it feel like a practical security research helper, so I can ask things like whether a domain has breach exposure, if a hash is known malware, or whether a target shows up in stealer logs and ransomware listings. Use current docs online if you need to, and make the setup as simple as possible, ideally something I can run with one command.