666ghj/MiroFish ? reverse-engineered prompt
Reverse engineered prompt
I want to build an AI prediction engine that acts like a digital sandbox for rehearsing the future. The idea is that a user can upload some source material—like a news report, a policy document, or even part of a novel—and then describe a scenario they want to explore.
The app would then create a simulated world filled with AI agents. These agents would have their own personalities and memories based on the initial documents and would interact with each other freely. By watching their collective behavior evolve, the system can predict future outcomes.
After the simulation is over, it should generate a detailed report summarizing the prediction. I also want an interactive frontend where I can explore this simulated world, almost like a game, and maybe even chat with the individual agents to understand why they made certain decisions. It should use an LLM for the agents' brains, so I'll provide an API key for that.
Are you gonna build this?
make sure you review the code using coderabbit