Agent Autoresearch

Karpathy-style autonomous self-research loop for AI agents. The agent proposes a change to its own SOUL.md, scripts, or behavior, tests it, evaluates the results, and either KEEP (integrates the change) or KILL (discards it). Runs continuously — experiments while you sleep. Works for any agent workspace. Triggers: "autonomous research", "self-improve", "run the experiment loop", "evolve yourself", "optimize my agent".

by @duolahypercho (Duola)·MIT-0