sota-code-security

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

• Third-party content exposure detected (high risk: 0.85). Skill runtime LLM context can be fed outsider-authored free text via the LLM/RAG/ingestion pipeline described in rules/08 (RAG chunks, scraped web pages, tool results) and rules/09 (scraped/webhook/uploaded content) reaching the model context for indirect prompt injection.