tech-article-extractor

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and analyzes user-generated content from open public sites (e.g., Qiita, Zenn, Stack Overflow, Reddit, Medium, YouTube, GitHub Discussions, arXiv), so the agent will read and interpret arbitrary third‑party content that could carry indirect prompt-injection.