learn-datalake

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

• Third-party content exposure detected (medium risk: 0.30). Source learn-datalake watches/ingests user-provided directory contents and feeds extracted requirement/passage text into the /match-requirement pipeline for autonomous requirement-to-control matching, so outsider-authored document text can reach the runtime LLM via the extraction→chunk→matching flow.