setup

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard filesystem commands (ls, cat, mkdir, chmod) to manage its own configuration directory and files within ~/.config/meigen. These operations are limited to the plugin's local state and are necessary for its intended functionality.
  • [EXTERNAL_DOWNLOADS]: Uses curl to perform connectivity tests to user-provided API endpoints (e.g., ComfyUI server or OpenAI-compatible services). These checks are used for validation purposes and do not involve downloading or executing arbitrary remote scripts.
  • [SAFE]: The skill implements security best practices by advising the agent to mask API keys (showing only the first 10 characters) and by explicitly setting file permissions to 600 on the configuration file to prevent access by other local users.
  • [SAFE]: Although the skill accepts a user-pasted curl command for easier configuration, it is instructed to parse the text for specific fields (API key, base URL) rather than executing the command directly, mitigating potential command injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 08:05 PM