AIPentest/CyberStrikeAI ? reverse-engineered prompt

Build me a CyberStrikeAI style web app for authorized security testing that lets me turn plain language requests into guided, auditable actions.

I want a dashboard where I can manage conversations, agents, roles, skills, tools, assets, vulnerabilities, and attack chain projects. The app should support human approval before risky actions, keep a full audit trail, save evidence, and let me replay what happened later. It should also include a knowledge base for security notes, a task queue for batch jobs, and a way to import and export assets from CSV or Excel.

Please make it feel like a real operations console, with a clean UI, dark mode, and status panels for runtime health and tool usage. If it helps, look up current docs online for anything uncertain.

Keep it safe by default, with access control and clear warnings around high risk features like remote shell or C2 style functionality, only for systems I own or am explicitly allowed to test.