39 citations · 39 across the 5 of their papers we have counts for
1 paper · 1 filter
Tony Mason
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 Arbit…