threadlinqs-cmd/intelthreadlinqs-mcp ? reverse-engineered prompt

Build me a TypeScript MCP server for threat intelligence like the one described here.

I want it to plug into Claude Code, Claude Desktop, Cursor, and any MCP client, and expose a large catalog of tools and prompt based analyst workflows for things like threat actors, IOCs, detections in Sigma, SPL, and KQL, MITRE attack chains, C2 infrastructure, CVEs, and correlation lookups. It should start cleanly from the command line with npx, read an API key from the environment, and handle access checks so the server can still list its tools even when the key is missing or not allowed, but actual data calls are denied in that case.

Please make the project feel production ready, with clear startup messaging, good error handling, and a simple way to add more tools and prompts later. If you need current MCP or Node docs, look them up online first.