“Every once in a while you go through your memories. If you guys have memory turned on, I do. If you go through your memory or if you go through CLAUDE.md, you can ask Claude, hey, which one of these maybe can turn into mods? And then this could be deleted from CLAUDE.md or memory.”

…those instead of relying on memory or clot.md instructions, you can just like bake it into your process. Yeah. And it just runs, right? So, I mean, like, I baked a lot of things based on these into the whole system. A couple of systems that I have, complex ones. Yeah, I think it works really great.

…even though I think cloud is also building this feature in a way that you know what the agent is doing, which one requires attention or not. But yeah, I think you from here you can just build your own agent to-do list that you can monitor easily rather than letting them just run randomly on your

Right. I mean, like one example that I'll give you is like, for example, like you know, I have this pipeline, right? So, in this pipeline, um, it would let's suppose I want to create an article, right? Let's suppose I have that pipeline of nine steps to create an article or a video, let's suppose…

Yeah, I feel like just you whenever you're talking, an idea popped in my head. If you go, you know, every once in a while you go through your memories. If you guys have memory turned on, I do. If you go through your memory or if you go through cloud.md and you can ask clothes, hey, which one of…

Yeah. Right. That's a cool idea. I'll try that. Yeah,