Skills
skills/kaerez/skills-plugins/puml-plantuml/Gen Agent Trust Hub

puml-plantuml

Fail

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The scripts/bootstrap.sh script downloads and executes the official Homebrew installation script using a shell command substitution pattern.
  • [COMMAND_EXECUTION]: The bootstrap and setup scripts (scripts/bootstrap.sh, scripts/bootstrap.ps1, scripts/setup.py) utilize elevated privileges via sudo to install system packages through platform-specific package managers such as apt, dnf, and pacman.
  • [COMMAND_EXECUTION]: The skill implements persistence mechanisms to ensure environment variables are available across sessions. It appends export commands to shell profiles (~/.bashrc, ~/.zshrc) on Unix-like systems and modifies user-level environment variables on Windows.
  • [COMMAND_EXECUTION]: The scripts/render.py utility invokes external processes using the Python subprocess module to manage diagram rendering and syntax validation tasks.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the plantuml.jar executable and other configuration data from official GitHub repositories during its automated setup process.
  • [PROMPT_INJECTION]: The skill processes external PlantUML files and images which represent an ingestion point for untrusted data. While the capability is limited to the PlantUML rendering engine, there is an inherent risk surface for indirect prompt injection if malicious instructions are embedded in the processed diagrams.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 17, 2026, 07:41 PM
Security Audit — agent-trust-hub — puml-plantuml