The Agent Skills Directory

[SAFE]: The skill utilizes internal platform libraries (omni.tracer and omni.foundation) to handle workflow compilation and execution tracing.
[SAFE]: File path construction for YAML pipelines in scripts/commands.py is restricted to the skill's own pipelines/ directory and validated against a predefined list of allowed pipeline types.
[SAFE]: No network communication, external data fetching, or hardcoded credentials were found in the scripts.
[SAFE]: The test_yaml_pipeline command uses a NoOpToolInvoker, ensuring that the pipeline compilation test does not trigger any actual tool execution or external LLM calls.
[SAFE]: The skill author ('tao3k') uses their own infrastructure and repositories as specified in the skill metadata, consistent with the provided vendor context.

demo