Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents

Bofan Chen , Boxuan Zhang , Fei Tang , Zhengxi Lu , Yong Du , Tongbo Chen , Weiming Lu , Jun Xiao , Yueting Zhuang , Yongliang Shen

🏛 Institutions: ZJU , UESTC
📅 Date: September 15, 2026
📑 Publisher: arXiv
💻 Env: Mobile Desktop
🔑 Keywords: framework EvoSkill-GUI skill evolution failure recovery training-free adaptation

TLDR

EvoSkill-GUI makes GUI agent skills revisable at deployment time instead of fixed beforehand, representing each skill as a multi-file package holding retrieval metadata, executable plans, backup localization, failure-recovery rules, and failure cases. A reflect-revise-reuse loop lets the executor patch skills mid-rollout while an isolated critic diagnoses failed trajectories, improving multiple base models without any training by up to 16.2% on MobileWorld, 6.0% on AndroidWorld, and 10.5% on OSWorld.

Open paper arXiv Code Report issue