hundred-million-leads

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly instructs scraping and ingesting public third-party sources (e.g., "Get a list -- scrape all contacts across platforms" in SKILL.md Warm Outreach and "social media scraping", "review site scraping", "Google Maps" in references/cold-outreach.md), meaning the agent would consume untrusted user-generated web content for personalization and outreach that could materially influence actions.