skills/moshe-ship/hurmoz/arabic-rag/Gen Agent Trust Hub

arabic-rag

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a standard Retrieval-Augmented Generation (RAG) workflow for Arabic text using established open-source tools.
  • [EXTERNAL_DOWNLOADS]: The skill downloads embedding models from Hugging Face Hub (AraGemma and Silma models), which is a well-known service for hosting machine learning resources.
  • [COMMAND_EXECUTION]: Instructions include the installation of official Python packages (sentence-transformers, faiss-cpu) via pip.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 07:15 AM
Security Audit — agent-trust-hub — arabic-rag