1 paper
Yihao Wang, Linhan Xia, Rui Liu +6
Large Language Model (LLM)-based agents increasingly self-evolve by editing a persistent skill document that encodes their workflow, tool-use rules, and decision logic. This loop h…