What problem does it solve?
Traditional userspace monitoring tools add performance overhead and miss kernel-level activity, leaving gaps in runtime visibility for Linux hosts and Kubernetes clusters. This Skill guides the deployment of eBPF-based monitoring with Cilium Tetragon so process execution, network connections, and file access are observed and enforced directly in the kernel.
Core Features & Use Cases
- TracingPolicy Authoring: Create kprobe and tracepoint policies with in-kernel matchArgs/matchBinaries selectors for file access, TCP connections, and privilege escalation detection.
- Runtime Enforcement: Apply Sigkill actions to block unauthorized binaries such as crypto miners at the kernel level, with guidance on staging policies in observe mode first.
- SIEM Integration: Export JSON events via file sinks or the tetra CLI into pipelines like Filebeat, Fluentd, or Elasticsearch.
- Use Case: A security engineer needs to detect reverse shells and container escape attempts across a Kubernetes fleet. Use this Skill to deploy Tetragon via Helm, apply TracingPolicies for tcp_connect and sys_openat, and stream matching events to Elasticsearch.
Quick Start
Ask the AI to deploy Tetragon on your Kubernetes cluster and create a TracingPolicy that monitors access to /etc/shadow and exports events as JSON.