The Agent Skills Directory

[SAFE]: The skill utilizes local helper scripts such as list_tao_platforms.py, list_tao_models.py, and resolve_tao_image.py to manage model and platform discovery. These operations are restricted to the local environment and are consistent with the skill's stated purpose.
[SAFE]: Data intake processes, including the collection of cloud platform credentials and dataset URIs (e.g., S3 paths), are performed by prompting the user for information. This is standard behavior for a deployment and training launcher.
[SAFE]: No malicious patterns such as obfuscation, remote code execution from untrusted sources, or data exfiltration to unauthorized domains were identified in the instructions or metadata.

tao-train-single-step