boost

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

• Third-party content exposure detected (low risk: 0.10). The workflow searches and fetches external repositories via git clone and skills.sh search, ingesting outsider-authored code and content from selected repositories.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill executes git clone using a hardcoded GitHub URL (https://github.com/{repository}.git) to dynamically fetch and install remote third-party source code at runtime, representing a weak/unknown/suspicious operational dependency.