RESEARCH

Towards Security-Auditable LLM Agents: A Unified Graph Representation

ArXiv cs.AI · Mon, 11 May 2026 04:00:00 GMT

arXiv:2605.06812v1 Announce Type: new Abstract: LLM-based agentic systems are rapidly evolving to perform complex autonomous tasks through dynamic tool invocation, stateful memory management, and multi-agent collaboration. However, this semantics-driven execution paradigm creates

Read original source Discuss with A.S.I.S