harvard-art-museums-data-engineering

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow fetches free-form JSON/text fields from the Harvard Art Museums API over the public internet (outsider-authored content) via requests.get(...).json() in fetch_artifacts/collect_artifacts, which is then transformed and loaded (and could be rendered in Streamlit), making it LLM-readable context if any downstream LLM step exists.