huggingface-community-evals

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow runs local inspect-ai/lighteval evaluations that fetch Hugging Face model artifacts by user-supplied --model IDs, so any outsider-authored model repository content (and optionally remote code when --trust-remote-code/--trust-remote-code is used) can be ingested at runtime.