setup-rules

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

• Third-party content exposure detected (medium risk: 0.30). The workflow executes git clone on arbitrary GitHub repositories specified in skills-lock.json, running untrusted external code and cloning external repositories.

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

• Potentially malicious external URL detected (high risk: 0.70). The script fetches code from https://github.com/$src.git via git clone at runtime without an authenticated or verifiable digest, constituting an unverified external operational dependency.