ai-cold-outreach

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.85). The document is a detailed, actionable playbook for scaling and evading detection in large‑scale cold outreach (domain/mailbox rotation, IP sharding, warmup circumvention, removal of traceable links, negative personalization) that can be used to support abusive mass‑spam or phishing campaigns, so it poses a high abuse risk.

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow ingests “recent company news” / “social posts or articles” and “research summary” for AI personalization, which are typically sourced from outsider-authored web content (e.g., G2/Bombora/LinkedIn pages or scraped articles) and then fed into the LLM via the AI personalization pipeline prompt.