priorart

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

• Third-party content exposure detected (high risk: 0.85). Step 3/4 require fetching and using fetched signals from outsider-authored public web/registry pages (e.g., GitHub/PyPI/Hugging Face/arXiv/Show HN/etc.) as part of the search “findings” that are then clustered into the LLM-rendered brief and verdict.