We’re deploying AI systems with access to tools… without real AI Security Posture Management

Hey everyone 👋

I built Orbix AI-SPM after realizing something uncomfortable:

👉 We’re deploying AI systems with access to tools and data… without real runtime control.

This project explores what a security control plane for AI could look like in practice.

Would love feedback on:

• Architecture decisions

• Missing attack vectors

• What would block you from using this in production

Happy to answer anything 🙏