pyvene-interventions

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides comprehensive guidance for using the pyvene library, an open-source framework developed by Stanford NLP for neural network interpretability research.
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of pyvene via official package registries and demonstrates loading pre-trained interventions from HuggingFace. The HuggingFace repository referenced (zhengxuanzenwu/intervenable_honest_llama2_chat_7B) is maintained by the primary author of the pyvene research paper, representing an official source for the library's research artifacts. These references to well-known services are informative and do not pose a security risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 07:53 PM
Security Audit — agent-trust-hub — pyvene-interventions