llm-agent-security

Here are 3 public repositories matching this topic...

Open admission layer for LLM agent harnesses: forbidden-path quarantine + cross-iteration escalation monitor. v0.1 heuristic, not full IFC.

Research-driven analysis of security architecture failures, trust boundaries, and attack paths in distributed and cloud-native systems.

Anonymous prepublication paper and reproducible artifact for EffectSvc-F.

To associate your repository with the llm-agent-security topic, visit your repo's landing page and select "manage topics."