The way I've done it has been when there's a longer task, I give it a markdown document that has a plan with numbered steps, and then for each step start a fresh context window and tell it to update that doc as it goes with any decisions taken or deviations from the plan, or other context needed for future steps. If it gets close to the end of the context I tell it to summarize the current state and anything a fresh context would need to know.
I was just looking at conductor and was not very jazzed about the fact that it was running the agents directly on the host. Being able to launch from a terminal means this one can (hopefully) run from inside the sandboxing setup I use for coding agents.
I think generally harnesses have this. In claude code it's `"preferredNotifChannel": "terminal_bell"` in the settings.json, pi and opencode looks like you have to either add a hook yourself somewhere or use an extension.
(Or computer mad scientist? Whatever (mad scientist) && (computer scientist) is)
reply