octave-research

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime path calls MCP enrichment/search tools (e.g., enrich_person, enrich_company, search_knowledge_base, find_person) that commonly fetch and ingest public web/community text about the target (outsider-authored free-form content) into the LLM context for summarization and output generation.