All Stories

  1. Introducing Safe LLM-Agent Execution via Typed Tool Mediation and World-Effect Evaluation

    Article • October 2026, ACM (Association for Computing Machinery)

    Jun Zhou