crypto-research

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

• Third-party content exposure detected (high risk: 0.80). The skill's News Scanner and agent prompts (see "news-scanner.md" and "Each agent prompt includes: Data gathering instructions (5+ tools)") explicitly describe aggregating recent news and external data, indicating the agents will fetch and read untrusted public third‑party content (news/social media/public websites) as part of their workflow.