skill-from-notebook

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 ingests public third‑party content (e.g., "Web URL" via WebFetch such as https://cbea.ms/git-commit/, public blogs like notes.eatonphil.com, GitHub repos, YouTube via yt-dlp, and NotebookLM pages) and is expected to read and interpret those user-generated/public web sources to extract methodologies, creating a clear avenue for indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches and ingests runtime content from user-supplied URLs (e.g., https://cbea.ms/git-commit/, https://notes.eatonphil.com/2024-04-10-what-makes-a-great-tech-blog.html, https://github.com/sindresorhus/awesome) to directly generate the skill's prompts/instructions, so those external URLs can control agent behavior at runtime.