geniml

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly references BBClient for accessing remote BED repositories and points to public sources like GitHub and Hugging Face for code and pre-trained models, meaning it can fetch and ingest open, third-party BED files and models (user-provided content) as part of its workflows.