@no7dw

openclaw-auto-training-skill runs autonomous qa evaluation loops, selects domain-specific tasks, scores agent responses with an llm judge, installs missing skills, and logs results to memory. triggers on user phrases like evaluate yourself or run qa.