AgentRuleBench: Detecting AI Agent Violations of Architecture Rules
AI · Studies · 2026-09-12 11:57:00
An open-source project that evaluates whether AI agents violate inferred architecture rules, aiming to improve their compliance and reliability in complex systems.