Building a Secure, Human-Approved Automated EC2 Alert Remediation System

Picture this: It's 3 AM. A critical alert fires because an EC2 instance is running out of memory or a process went rogue. Traditionally, an engineer wakes up, finds their laptop, logs in, SSHs into th