sue-runpod-recover

What problem does it solve?

When a RunPod pod reboots, all user Linux accounts, SSH login access, and pre-installed development environments are wiped, leaving operators unable to access their persistent workspaces and requiring hours of manual reconfiguration of user accounts, SSH keys, and tooling.

Core Features & Use Cases

• Automatic User Account Restoration: Recreates Linux user accounts with persistent /workspace home directories after reboot, preserving all stored experimental data and artifacts.
• Seamless SSH Auto-Login: Reinstalls pre-captured user SSH public keys to enable passwordless login without manual re-keying, restoring immediate access to workspaces.
• Development Environment Recovery: Reinstalls conda, pip, and apt packages captured during pre-reboot precheck to restore development environments to their exact pre-reboot state.
• Use Case: If your RunPod pod reboots mid-ML experiment and you cannot log in to your workspace, this skill restores all user access and tooling in minutes instead of hours of manual setup.

Quick Start

Use the sue-runpod-recover skill to restore all RunPod user accounts, SSH login access, and development environments immediately after a pod reboot using the pre-captured precheck bundle.