huggingface-vision-trainer

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow explicitly fetches and parses public Hugging Face Hub datasets (e.g., calls to load_dataset("username/dataset-name") and the dataset_inspector.py use of the Datasets Server API URL https://datasets-server.huggingface.co/rows?...), which are untrusted user-generated sources and whose contents are interpreted to decide preprocessing, training configuration, and job submission—so third-party content can materially influence agent actions.