pathogen-variant-surveillance

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

• Third-party content exposure detected (low risk: 0.10). The skill queries public LAPIS genomic databases and fetches Pango designation files from GitHub, which are community and provider registries containing outsider-submitted text.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill fetches configuration data from GitHub during execution, and since third-party user-controlled repositories are classified as weak or unknown provenance sources, this constitutes an unverifiable runtime external dependency under the high tier.