🧰 Mine

The items that name it 2

• [tl;dr sec] #352 - Unprompted Slides, AWS AI Post-compromise Trends, Malicious LLM API Routers
  article2026-10-29

  Research proxy from the Hanzhi Liu et al paper that implements all four attack classes against four public agent frameworks, used to evaluate client-side defenses.

• Your Agent Is Mine: Measuring Malicious Intermediary Attacks on the LLM Supply Chain
  paper

  Research proxy built by the authors implementing all four attack classes (AC-1, AC-1.a, AC-1.b, AC-2) against four public agent frameworks, used to evaluate client-side defenses