Research · curated 30 Aug 2026
Agent Against Agent: An Agentic System for Automatic Prompt Injection Red Teaming
First reported arxiv.org
Coverage timeline
Single-source research — first reported, latest, and curated coincide.
Why it matters
PIMiner demonstrates that a learned, transferable prompt-injection attack library can achieve high success rates against frontier LLM agents with minimal queries, giving defenders both a benchmark of agent vulnerability and a data-generation pipeline for hardening guardrails.
PIMiner, described in an arXiv paper by researchers at Pennsylvania State University, is an agentic system for automatic prompt injection red-teaming that builds a transferable strategy library and attacks unseen target LLM agents with only a few queries per sample. It reports high attack success rates against agents including Gemini-2.5-Pro (76.2%/86.7% on IPIArena/AgentDojo), GPT-5.1, and Claude-Sonnet-4.5, with code released at github.com/wang-yanting/PIMiner.