alterlab-imaging-data-commons

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to query and ingest public IDC datasets (e.g., via idc-index client.sql_query(), client.fetch_index(), BigQuery guides, and DICOMweb/public S3/GCS URLs such as the DICOMweb proxy and series_aws_url) and then use those results to drive downloads, viewer URLs, and processing decisions, which clearly exposes the agent to untrusted third-party content that can influence actions.