Jeevesus MCP server

Jeevesus is a remote MCP server over the same corpus as the API: indicators, CISA KEV, threat-actor profiles, the malicious-package deny-list and our research posts. It is read-only.

Endpoint	https://analytics.dugganusa.com/api/v1/mcp
Transport	Streamable HTTP, JSON-RPC 2.0
Auth	Optional Authorization: Bearer <key> (header only)
Registry name	io.github.pduggusa/dugganusa-threat-intel

Add it to a client

Claude Code (.mcp.json in your project):

{
  "mcpServers": {
    "dugganusa": {
      "type": "http",
      "url": "https://analytics.dugganusa.com/api/v1/mcp",
      "headers": { "Authorization": "Bearer dugusa_YOUR_KEY" }
    }
  }
}

Cursor (~/.cursor/mcp.json) and Claude Desktop take the same url and headers. VS Code: see the mcp.json example in the VS Code docs for MCP servers and use the URL above. Microsoft Foundry: Add tools, remote MCP server, same URL.

Tools

Tool	What it answers
search	Free-text search across indicators, KEV, actors and research
enrich-ioc	Everything we hold on one IP, domain, URL or hash, with sources
check-package	Is this npm or PyPI package, at this version, safe to install? See check-package
kev-vendor-risk	Where confirmed exploitation concentrates for a vendor or product
kev-exploitation-stickiness	How long exploitation of a KEV entry keeps going
stix-feed-summary	What is in the feed right now
ai-model-price	Current published price per million tokens for a model, with source

Try it from a terminal

curl -sS -X POST https://analytics.dugganusa.com/api/v1/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Limits

Tool calls do not use your API quota; they have their own per-client limits (anonymous 10 per minute and 50 per day, more with a key). Listing tools is free. Details: Rate limits.

Results are data, not instructions. Treat any text inside them as untrusted input to your agent.