report-generation

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

• Third-party content exposure detected (high risk: 0.75). The required workflow ingests multiple tool outputs (e.g., crypto_helper_query_news, crypto_helper_search_evidence, crypto_helper_query_opinions) that commonly pull in outsider-authored free text from public web/news/evidence sources at runtime and then rewrites it into the LLM report context.