chanchalvdev/go-siem-agent-llm-classifier ? reverse-engineered prompt

Reverse engineered prompt

Build me an open source SIEM style security app in Go with a React dashboard.

It should take in security logs from syslog, pasted text, or uploaded files, turn them into structured events, detect obvious attacks with rules right away, and use an LLM to classify the rest with a severity score, confidence, and MITRE ATT&CK mapping. I want incidents to be grouped from related alerts, with a timeline, comments, assignee, status, and a way to export a report. It should also pull out useful indicators like IPs, domains, and hashes, keep a searchable history, and show live updates in the UI.

Please make it work with Docker so I can run a full demo easily, and include basic auth with roles, audit logs, and some simple response actions like blocking an IP or disabling a user after approval. If you need to check current docs for any integrations, go ahead and look them up online.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab