Design and execute infrastructure-level chaos experiments targeting nodes, disks, CPU, memory, network, and cloud resources. Use this skill whenever the user says "test node failure", "drain a node", "fill disk", "test CPU exhaustion", "simulate AZ failure", "stop EC2 instances", or needs infrastructure-level fault injection manifests for LitmusChaos or Harness Chaos Engineering. Also trigger as part of the chaos experiment design phase (s16 follows s12).
