trl-training

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's CLI and docs explicitly instruct loading datasets from public sources (e.g., dataset_name like "trl-lib/Capybara" and the advice "Inspect dataset: from datasets import load_dataset; ds = load_dataset(name)"), meaning the agent workflow ingests untrusted, user-hosted Hugging Face datasets which can materially influence training and tool behavior.