Hi - I am interested in /goal, but the documentation doesn't tell me what the internal algorithm is. I'd like to know the steps it takes, where it stores state, how it fires subgoals, what information is provided to subgoals, etc.. Even better, I'd like to build on this infrastructure (which is great!0. If anyone could give me a pointer, I'd be happy to report back on things I try. In particular, I am looking for exactly the opposite of what the documentation says /goal is good for - I am interested in high-level, long-term, autonomous processing of broad goals.

Are you talking about this documentation? https://docs.openhands.dev/sdk/guides/convo-goal

You can also point your agent at agent-sdk to find it, understand it, make you a visualization or ask it questions. It’s open source 😀

Beautiful, thanks, I missed that document somehow.

Actually, the goal alg is very basic for now. I have some ideas on how to improve it, but every advice is welcome :)