paoding

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow ingests outsider-authored blog posts, video transcripts, and comments via user-provided text, links, and automated download scripts (yt-dlp and API clients).

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

• Potentially malicious external URL detected (high risk: 0.70). The skill contains operational Python and shell scripts that query and download content from third-party APIs and websites (TikHub, SocialData, and social media platforms via yt-dlp) without enforced signatures or official digests, constituting a weak runtime external dependency.