parity-testing

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

• Third-party content exposure detected (high risk: 0.80). This skill's SKILL.md explicitly instructs running conversion and comparison scripts (examples/conversion/*) with --hf-model-id / --hf_model_path like /, which loads models and their generated outputs from the public Hugging Face hub (untrusted third-party content) and uses those artifacts as inputs that affect parity checks and next debugging actions.