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's SKILL.md and reference guides explicitly direct the agent to query and ingest public third‑party data (e.g., "File downloads use public GCS (storage.googleapis.com) and AWS S3", DICOMweb proxy, and BigQuery public datasets such as bigquery-public-data.idc_current) and to consume manifest files/portal exports (USAGE.md and cloud_storage_guide.md), so the agent will read and act on untrusted, user-provided public content that can materially change tool actions (what to download, view, or analyze).

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

• Potentially malicious external URL detected (high risk: 0.80). The skill includes a runtime package upgrade that will fetch and install the required idc-index package (e.g., via pip from https://pypi.org/project/idc-index/ or the GitHub repo https://github.com/ImagingDataCommons/idc-index), which downloads and executes remote code that the skill relies on.