Reduce LLM token consumption by 60-90% by filtering CLI output through snip

Install

openclaw plugins install clawhub:openclaw-snip