xalgorix/xalgorix ? reverse-engineered prompt

Build me a self hosted AI pentesting app that I can run locally on Linux or in Docker. It should have a web dashboard and a simple CLI, let me connect my own LLM provider or use local models, and guide me through setup the first time.

I want it to automatically do recon, run common security checks, look for likely vulnerabilities, and then try to prove the issue before it reports it, so I get real findings instead of a bunch of false alarms. It should keep a clear record of what it tested, what it found, and how confident it is, and let me review results in the browser.

Please make it feel like a real security tool for authorized testing, with sane defaults, persistent settings, and easy start commands. If you need to check current docs for any tools or model setup details, look them up online first.