guardian-cli-ai-pentest

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions include cloning a repository from GitHub (https://github.com/zakirkun/guardian-cli.git) and installing it locally. The source is a public repository, and this is standard for open-source security tooling.
  • [COMMAND_EXECUTION]: The skill provides numerous examples of executing CLI commands (python -m cli.main ...) and shell commands for installation and configuration (pip install, git clone, export). These are intended for local setup and operation by the user.
  • [CREDENTIALS_UNSAFE]: The documentation includes placeholders for API keys (sk-your-key-here, your-gemini-key, ${OPENAI_API_KEY}). It correctly advises users to manage secrets via environment variables or a configuration file (config/guardian.yaml), which is an accepted practice for developer tools. No actual secrets are hardcoded in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 11:44 PM