OpenClaw
- Source
- lexfridman.com
- Date
Context
OpenClaw's proactive messages come from a recurring prompt. Its creator, Peter Steinberger, says in the interview transcript that the first version was simply a prompt, repeated about every half hour, telling the to surprise him. He later made that instruction more specific.
The feature, discussed as Heartbeat, lets the assistant act without waiting for an incoming message. Steinberger describes it as a follow-on to the current session and, in the end, a cron job (a scheduled task) that kicks off the loop. Because it continues the existing session, the agent can draw on the conversation already underway and sometimes asks a follow-up question. The design choice that matters is when the schedule fires and what it asks the agent to do.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
- agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.
Checking sign-in…
Loading comments…