pp-wanderlust-goat

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

• Third-party content exposure detected (medium risk: 0.30). The skill workflow involves syncing and fetching data from third-party social sources like Reddit and public travel platforms (via sync-city, reddit-quotes, and near), which ingest outsider-authored free text into the agent's runtime.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the installation of external Go and Node packages from GitHub and npm repositories (github.com/mvanhorn/printing-press-library/... and @mvanhorn/printing-press-library), which constitute unverifiable runtime external dependencies.