pull down to refresh
Heh. One of the things I was worried most about was this regular heartbeat thing. Don't want any surprise bills
I put a self-assessment prompt in there and it started burning millions of tokens (on my local model.) So I told it to stop doing anything at all... and then shut it down. Waste of precious GPU ticks, battery life and otherwise watt hours... I don't have anything for it to do anyway.
Wait but if you write HEARTBEAT is disabled to a new HEARTBEAT.md, won't it still run the LLM on that? Not sure how the heartbeat system works, I just know there is no intelligence until it calls the LLM, prior to that it's all just rules.
Yes it will, but then it will do nothing-ish unless you pass it into Qwen3.5, which will do 6 hours of hypothesis in reasoning mode about what that could possibly mean but wait maybe it means something else.
Can disable it in settings, but I don't let a bot into its own settings. And it was easier to just turn it off.
And the superprompt does seem to work pretty well to get the bot to be self-actualizing.
For example, if I ask regular Claude to tell me how to edit the openclaw config files to do X, it won't do as good of a job as if I ask the clawbot to edit its own files to do X.
That self-sufficiency aspect of knowing how to operate itself is pretty interesting. It's like the brain is actually connected to the hands and feet, so to speak.