The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill clones the i4h-workflows repository from https://github.com/isaac-for-healthcare/i4h-workflows. This is a trusted resource owned by the skill's authoring team at NVIDIA and is necessary for the skill's primary function of robotic policy training.- [COMMAND_EXECUTION]: The skill uses standard shell commands to manage directories, check GPU status (nvidia-smi), and execute training CLI tools within a controlled environment (uv run). These operations are scoped to the skill's purpose and do not exhibit privilege escalation or persistence patterns.- [DATA_EXFILTRATION]: No sensitive data access or unauthorized network exfiltration was detected. The network operations are limited to cloning the official repository and reporting training metrics to well-known services like TensorBoard or Weights & Biases if explicitly configured by the user.- [PROMPT_INJECTION]: The instructions are task-oriented and do not contain patterns typical of prompt injection or safety guideline bypasses. The skill uses clear delimiters and structured workflows.

i4h-workflow-finetune