torch-geometric

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

• Third-party content exposure detected (high risk: 0.80). The skill's "Loading Built-in Datasets" and training workflow explicitly show loading public datasets (notably "Reddit" and other Planetoid datasets) which are open, user-generated third‑party content that the agent ingests and uses to drive model training/decisions, exposing it to potential indirect prompt injection.