reply-audit

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

• Third-party content exposure detected (high risk: 0.95). SKILL.md Step 0 ingests outsider-authored free text from sources like “Google Sheet/Docs link”, “Notion page/table URL”, “Outreach-platform campaign… pull conversations/replies”, or “Raw pasted text”, which then becomes LLM-readable reply content used in Steps 1–4 (classification, template extraction, and per-lead ready replies).