repurpose-discord

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

• Third-party content exposure detected (medium risk: 0.65). The skill ingests atoms/main_argument/etc. from the parent agent (repurpose-community); if those fields include outsider-authored free text (e.g., scraped web content, issue/wiki/discussion text, or other users’ messages) then that prose is incorporated into the generated Discord outputs and thus into the LLM context at runtime.