skills/openai/plugins/transformers-js/Gen Agent Trust Hub

transformers-js

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • External Resource Integration: The skill fetches pre-trained models from the Hugging Face Hub and loads library components from the JSDelivr CDN. These integrations are standard for machine learning workflows, allowing applications to access large model weights and runtime dependencies without local bundling.
  • Execution Environment Isolation: Machine learning inference is performed using WebAssembly (WASM) or WebGPU. This provides a high-performance execution layer that operates within the security sandbox of the JavaScript runtime, isolating model computation from the host system.
  • Local Data Caching: The skill manages model files by storing them in local cache directories (either via the filesystem in Node.js or the Cache API in browsers). These operations are used to optimize performance on subsequent loads and minimize external network requests.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 05:44 PM
Security Audit — agent-trust-hub — transformers-js