repo-wiki

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

• Third-party content exposure detected (medium risk: 0.30). In references/extraction.md, the chat extraction subagent takes a transcript read from scripts/vendor/recall/read_session.py <file> as an input and uses Prompt 1 to propose knowledge pages from that content, so outsider-authored chat text can be injected via the chat/recall transcript intake path.