anxcrn/ravanalex ? reverse-engineered prompt

Build me a security research and training app inspired by this repo, but keep it clearly safe and for authorized lab use only. I want an AI driven red team simulator that feels like a swarm of specialist agents working together to analyze a target environment, map likely attack paths, relate findings to common security frameworks, and generate clear reports and remediation advice. It should run locally, let me plug in one of the major LLM API keys, and have a polished interface where I can start a simulated assessment, watch the agents coordinate, and review results.

Please do not include anything that performs real exploitation, credential theft, phishing, malware creation, persistence, exfiltration, or stealth behavior. Use mock targets, sandboxed workflows, and sample data so it is useful for demos, education, and defensive planning. If the repo already has the right structure, wire it up into a working end to end experience and add setup steps so I can run it easily. You can check current docs online if needed.