linkfox-zhihuiya-abstract-data-translated

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/zhihuiya_abstract_translated.py, the runtime accepts user-supplied JSON (patentId/patentNumber/lang) and then ingests the provider response fields (title/abstractText) returned by POST /zhihuiya/abstractDataTranslated into serialized output/stdout without filtering, so outsider-authored free text embedded in those returned records becomes readable to the LLM workflow.