classify-technical-evidence-ip

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

• Third-party content exposure detected (medium risk: 0.30). The skill ingests user-supplied spreadsheet/CSV “Source Data” text and, if external enrichment is authorized, uses that text as part of the “Execute the authorized discovery profile” (keyword expansion/semantic/field retrieval/focused patent or literature retrieval) before classification, making outsider-authored free text able to influence runtime LLM processing.