model-verify-flagos

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

• Third-party content exposure detected (low risk: 0.10). The workflow downloads and runs models from HuggingFace/ModelScope repositories, which can contain untrusted code/metadata executed inside the container environment.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill downloads model weights from ModelScope at runtime using snapshot_download('<MODEL_ID>'), which constitutes an unverified external data/code dependency.