torch-geometric

Pass

Audited by Gen Agent Trust Hub on Oct 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill follows established patterns for PyTorch Geometric usage and installation from official sources.
  • [EXTERNAL_DOWNLOADS]: Fetches library components and datasets from official domains like data.pyg.org and pytorch.org.
  • [INDIRECT_PROMPT_INJECTION]: The skill implements data loading patterns that ingest external graph and text data.
  • Ingestion points: custom CSV loaders and Planetoid/TUDataset datasets.
  • Boundary markers: Not present.
  • Capability inventory: Network downloads, file system writes for data caching, and model execution.
  • Sanitization: Data is transformed into numerical tensors, which naturally limits typical text-based injection, though no explicit string sanitization is demonstrated.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 1, 2026, 07:50 AM
Security Audit — agent-trust-hub — torch-geometric