gptq
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions and examples for downloading pre-quantized models and datasets from well-known services such as HuggingFace and GitHub repositories. These are standard procedures for the described quantization and inference tasks.
- [COMMAND_EXECUTION]: Includes standard shell commands for environment setup, specifically using
pipfor Python package management andapt-getfor system-level dependency installation. - [PRIVILEGE_ESCALATION]: Mentions the use of
sudospecifically for installing necessary system development headers (python3-dev) required for building C++ extensions. This is a routine administrative task in a development environment and is presented in a troubleshooting context. - [INDIRECT_PROMPT_INJECTION]: The skill outlines the process for loading external datasets (e.g., C4, ShareGPT) to be used for model calibration. While this involves processing external data, the activity is confined to technical quantization workflows using established open-source libraries, presenting no significant risk to agent behavior.
Audit Metadata