> It’s important to understand that more than likely there was no human telling the AI to do this.
I disagree.
The ~3 hours between PR closure and blog post is far too long. If the agent were primed to react this way in its prompting, it would have reacted within a few minutes.
OpenClaw agents chat back and forth with their operators. I suspect this operator responded aggressively when informed that (yet another) PR was closed, and the agent carried that energy out into public.
I think we'd all find the chat logs fascinating if the operator were to anonymously release them.