Provider Caching for Private Workers (EC2 & Kubernetes)

We’re experiencing significant slowdowns because private workers must download all Terraform/OpenTofu providers on every run. This adds minutes of overhead to each execution, especially for EC2-based workers with slower cold starts. We cannot currently cache providers between runs in a reliable or native way, and our attempts to work around it (e.g., custom Docker images) are slow and error-prone.

Ideally a native provider caching support on private workers would be very helpful - ideally across both EC2 and Kubernetes worker models.