terechgracz wrote:Nice stuff. Are you planning to add LLM support? I can see this client being useful for debugging or building an LLM-based agent system to control hearthlings. This approach wouldn't be ideal for actual botting, but it could be a great tool for educational purposes.
I wish I could tell a hearthling to go and kill snail... jokes aside pretty much everything I would be able to tell it to do would have to rely on a precoded set of actions that the model could refer to, and thats my current aim (routing engine, managing inventory, automatic routine tasks, etc...).
I think integrating an LLM model might come handy for a large hive mind, otherwise a simple UI will do the trick. Lets just hope the motivation lasts.
noindyfikator wrote:Are you going to open source the project? More wizards can contribute
At the moment I really want to become the coolest kid in the playground.
