scene-asset

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

• Third-party content exposure detected (low risk: 0.10). The referenced GitHub documentation files (references/cinematic-image-direction.md) instruct the agent to read external repository links (github.com/freestylefly/...), which exposes the agent to outsider-authored text when followed.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill references GitHub URLs for style galleries and templates (https://github.com/freestylefly/awesome-gpt-image-2/...) under an individual's unverified third-party repository, constituting a weak/unknown/suspicious operational dependency.