AI RESEARCH
Arbiter: Detecting Interference in LLM Agent System Prompts
arXiv CS.AI
•
ArXi:2603.08993v1 Announce Type: cross System prompts for LLM-based coding agents are software artifacts that govern agent behavior, yet lack the testing infrastructure applied to conventional software. We present Arbiter, a framework combining formal evaluation rules with multi-model LLM scouring to detect interference patterns in system prompts.