Open-source runtime AI agent security tool - monitors and controls AI agents, catching malicious tool use, prompt injection, and policy drift in real time, before the agent acts.
ai mcp self-hosted agents observability ai-agents ai-security threat-detection runtime-security ai-governance langchain prompt-injection llm-security langgraph agentic-ai agent-security openai-agents mcp-security aarm agentic-security
-
Updated
Sep 15, 2026 - Python