dataset-discovery

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

• Third-party content exposure detected (high risk: 0.90). This skill programmatically fetches and ingests untrusted, user-generated content from public sites (e.g., HuggingFace API in search_huggingface and cmd_pull via datasets-server endpoints, OpenML API in search_openml, GitHub via gh in search_github and cmd_detail which reads repo readme, and Semantic Scholar in search_papers), and that fetched text and metadata are parsed and used to rank/deduplicate results and drive follow-up actions (detail/pull), so third-party content can materially influence agent behavior.