to-artifact

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

• Third-party content exposure detected (medium risk: 0.30). The workflow mentions taking a subject from a PR or issue related to the current task if specified, which introduces exposure to untrusted outsider-authored text via GitHub issues or PRs.

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

• Potentially malicious external URL detected (medium risk: 0.30). The stylesheet dynamically loads external Google Fonts via CSS <link> tags at runtime, representing an inbound operational dependency from an established third-party provider, placing it at the medium tier.