blip-2-vision-language

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill is an educational resource providing documentation and Python code snippets for multimodal AI tasks. No malicious behaviors, obfuscation, or unauthorized data access patterns were identified.
  • [EXTERNAL_DOWNLOADS]: The instructions direct users to install official libraries from PyPI (e.g., transformers, salesforce-lavis, torch) and download pre-trained weights from the Hugging Face Model Hub. These sources are well-known, industry-standard services and the organizations (Salesforce, Hugging Face) are trusted entities in the AI ecosystem.
  • [COMMAND_EXECUTION]: The documentation includes standard shell commands for package management (pip install) and web server execution (uvicorn). These are typical and expected for the development and deployment of machine learning applications.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 12:55 AM
Security Audit — agent-trust-hub — blip-2-vision-language