newsletter-digest

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

• Third-party content exposure detected (high risk: 0.90). The skill's fetch-feeds.js reads arbitrary public RSS feed URLs from feeds.json (e.g., Hacker News, The Hacker News, etc.), writes article content to /tmp/newsletter-digest-articles.json, and SKILL.md Step 4 then embeds those fetched article summaries directly into the Gemini synthesis prompt — meaning untrusted, third‑party content is ingested and used to drive model decisions and outputs.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime fetch-feeds.js reads feeds.json and pulls article content from external RSS URLs (e.g., https://feeds.feedburner.com/TheHackersNews, https://hnrss.org/frontpage, https://www.infoq.com/feed/, https://changelog.com/feed, https://css-tricks.com/feed), and that fetched content is injected into the Gemini prompt as the sole sources, so remote content can directly control model instructions.