money-outreach

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

• Third-party content exposure detected (high risk: 1.00). The skill's Phase 1 research and Phase 2 list-building explicitly require finding and using untrusted public content via "Web search for companies matching ICP" and named third-party sources (LinkedIn, G2/Capterra, GitHub, Twitter/X) and extracting personalization hooks (recent post, company news) that the agent must read and act on.