Obsidian Security Launches Runtime Policy Engine for AI Agent Actions

Obsidian Security's new Runtime Policy Decision Engine evaluates AI agent actions before execution, adding an MCP Security Proxy, Policy Decision API, and OpenAI Codex support.