tao-run-automl

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly requires and uses an external LLM endpoint at runtime for LLM-guided algorithms (default https://inference-api.nvidia.com), and calls to that endpoint return LLM responses which directly drive hyperparameter proposals/agent decisions — i.e., remote content from that URL can control prompts/behavior during execution.