kortix-ai/suna ? reverse-engineered prompt

I want to build a platform that's like an operating system for AI agents, something you could call an "autonomous company OS." The core idea is to have a system where AI agents can work for me 24/7. Each agent needs its own secure Linux sandbox environment where it can run commands, write code, manage files, and interact with APIs, just like a person would.

They need to have persistent memory so they can remember context and learn over time. I also want to be able to trigger them automatically, either on a set schedule like a cron job, or in response to external events using webhooks. Finally, the whole thing should be easy to install on a server with a simple one-line script, and it should come with a basic command-line tool to let me start, stop, and manage the system.