social-intelligence

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md instructs the agent to call enrichx402.com endpoints (e.g., https://enrichx402.com/api/grok/x-search and https://enrichx402.com/api/reddit/search) to fetch X/Twitter posts and Reddit posts/comments, which are untrusted, user-generated third‑party content that the workflows (Brand Monitoring, Community Sentiment, Competitor Research) require the agent to read, summarize, and base decisions on.