Onyx Security provides an AI agent security platform for enterprise security teams. The platform discovers AI agents across the organization, builds an inventory, and continuously assesses how those agents access systems and data. Security teams define policies for what agents are allowed to see and do, and Onyx enforces these rules with real-time guardrails. The Onyx Guardian Agent monitors agent behavior at runtime, flags excessive permissions, blocks unsafe actions, and helps incident response when something goes wrong. The goal is to let enterprises adopt agentic AI at scale without losing visibility, control, or compliance.