peft-fine-tuning

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 loads public Hugging Face models and datasets (e.g., load_dataset("databricks/databricks-dolly-15k") and AutoModelForCausalLM.from_pretrained("meta-llama/...") and links to huggingface.co/github), which are open, potentially user-generated third-party content that the workflow ingests and uses to train or modify model behavior.