One Container, Five Risks: How Wiz's Green Agent Maps the Fastest Way to Break an Attack Chain
Wiz Research has outlined how its Green Agent tool handles a critical scenario: an internet-facing container running an image with a known initial-access vulnerability, with a role attached that can reach sensitive data. With AI reducing the time from discovery to exploitation, the pressure to fix such problems quickly is intense. Yet teams must also protect business continuity, and they need to work out what is impacted, what the fix is, and who should receive it.
In the example, the Wiz Security Graph mapped a single Kubernetes deployment in an AWS EKS cluster where five risks converged. These were external exposure from the internet to the container, a known vulnerability (CVE-2026-111497) on the image, a pod mounting a service account token that bridges into the cloud, an IAM role granting access to cloud storage, and sensitive customer records held in the accessible buckets. The graph shows the path an attacker could follow from code execution to storage holding personal information.
Wiz notes that finding the fix and the right owner takes deeper investigation than security teams can afford to spend hours on. Green Agent is designed to do this in minutes. The five risks also mean five places to break the chain: remove the public exposure, patch the container image, unmount the service account token, scope down the IAM role, or purge the sensitive data. This is only the opening of the Wiz post, so further detail on its recommendations is not covered here.