news-aggregator-skill

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

• Third-party content exposure detected (high risk: 1.00). The skill (SKILL.md and scripts/fetch_news.py) explicitly scrapes and fetches full article text from open/public third-party sources (Hacker News, Weibo, V2EX, GitHub Trending, 36Kr, Tencent, WallStreetCN, Product Hunt) using a --deep mode and requires the agent to read and deeply analyze that untrusted, user-generated public content to drive report generation and prioritization, which could allow indirect prompt injection.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill's scripts (scripts/fetch_news.py) perform runtime HTTP fetches of arbitrary article URLs via fetch_url_content (requests.get on item['url'], e.g. links from https://news.ycombinator.com or other external article URLs) and then inject the retrieved text into item['content'] which is subsequently used for the required "deep analysis" prompts—meaning remote content fetched at runtime can directly influence the agent's model input.