Abdul-Haseeb-sec/Omnilog ? reverse-engineered prompt

Build me a security analysis app that lets me upload raw logs or packet captures and then shows me suspicious activity in a clean dashboard. It should accept common formats like pcap, pcapng, cap, log, csv, json, and gz, then automatically figure out what the file is and run detections for things like SSH brute force, DNS anomalies, HTTP errors, port scanning, and basic data exfiltration patterns. I want a web dashboard where I can click into each alert, see the details and timeline, filter by true positives, false positives, and needs review, and tag IPs as safe or malicious so those tags are remembered next time. Also add real threat intel lookups when possible, using a free source if no key is set. Please make it work locally first, with a backend API and a simple frontend, and keep it easy to run with Docker for production.