large-file-refactor
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for local source code analysis and refactoring planning. It uses a Python helper script to scan directories and identify large files.
- [SAFE]: Command execution within the Python script utilizes list-based arguments in
subprocess.runand sanitizes user-provided arguments usingshlex.split, which effectively prevents command injection vulnerabilities. - [SAFE]: The skill implements a 'human-in-the-loop' safeguard by explicitly instructing the agent not to execute any refactoring steps unless the user provides a separate, explicit request.
- [SAFE]: External dependencies are managed through standard tools (
uv) and targeted at well-known development utilities (git, tokei, rg) already present in many developer environments.
Audit Metadata