Tired of checking saved RunPod GPU pods, a solo dev built a Chrome extension

A practitioner who reuses saved GPU servers (pods) on RunPod ran into a recurring annoyance. Their saved pods already have a fully working setup — CUDA/PyTorch sorted, repositories cloned, packages installed, and models, configs, and checkpoints downloaded from earlier runs. But when trying to restart a pod, the GPU was often unavailable, forcing constant manual refreshing to check when it freed up.

RunPod's built-in Cloud Sync only moves files around; it doesn't restore the entire pre-configured working environment, so it wasn't a real fix. To solve this, they built PodScout, a that monitors saved pod availability, sends , and can automatically claim a saved pod the moment it becomes available. It also offers optional 24/7 monitoring with SMS alerts for when the user is away from their laptop.

The extension recently launched on the , and the developer is now gathering feedback from other RunPod users to gauge interest and improve it.

Key points

  • Saved RunPod pods retain a fully configured environment (CUDA/PyTorch, repos, packages, models) but restarts often fail due to GPU unavailability
  • RunPod's Cloud Sync only moves files, not the full working environment
  • PodScout monitors saved pod availability, sends , and can autoclaim a pod when it frees up
  • Optional 24/7 monitoring and SMS alerts are available for when the user is away
  • Now live on the , with the developer collecting user feedback
Read original