fine-tuning-expert

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

• Third-party content exposure detected (high risk: 0.90). The skill's dataset and deployment references (e.g., references/dataset-preparation.md's load_custom_dataset which calls load_dataset(path) and the deployment/quantization guide using load_dataset("wikitext", ...)) explicitly load public/Hugging Face datasets and arbitrary dataset paths, so the agent ingests untrusted third-party/user-generated content that can directly influence training, calibration, and downstream model behavior.