The core concept of agent graphs seems quite similar to an LLM agent framework I created last year, [1]. Imo it is a more promising way to make agents, than th

The core concept of agent graphs seems quite similar to an LLM agent framework I created last year, [1]. Imo it is a more promising way to make agents, than the prior approach of just giving an LLM N tools to use at once. [1] https://github.com/sradc/MakeAgents https://github.com/sradc/MakeAgents