The Agent Skills Directory

[Command Execution]: The scripts/mirror_image.py script utilizes the subprocess module to execute aws and docker commands. The implementation follows security best practices by passing arguments as lists and avoiding the use of shell=True, which significantly reduces the risk of shell injection.
[External Content Access]: The skill fetches container image information and model configurations from official AWS and Hugging Face repositories. These are well-known, trusted sources necessary for determining deployment compatibility and mirroring images to private registries.
[Indirect Prompt Injection Surface]: The workflow involves querying config.json files from the Hugging Face Hub to identify model architectures. While this represents an ingestion of external data, the content is sourced from a trusted platform and used to inform manual or scripted deployment decisions, presenting a minimal security consideration.

hf-cloud-serving-image-selection