lamindb

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests public third-party content — e.g., importing public ontologies via bt.CellType.import_source() and loading data from HTTP/HTTPS endpoints and HuggingFace datasets as shown in SKILL.md (see Integrations and the ontology use-case), so untrusted external content is read and used to influence processing and annotation.