setup-wizard
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple shell commands to initialize a Laravel project, manage directories, and configure local server settings using tools like the Laravel installer, Composer, and Laravel Herd.
- [INDIRECT_PROMPT_INJECTION]: User-provided input for the project name and directory is interpolated directly into shell commands, creating a potential command injection surface. * Ingestion points: User responses to prompts for PROJECT_NAME and PROJECTS_DIR in SKILL.md. * Boundary markers: The agent is instructed to enforce naming rules (lowercase, numbers, and hyphens only) for the project name, which serves as a natural language boundary. * Capability inventory: The skill uses mkdir, cd, laravel new, composer require, and php artisan across SKILL.md. * Sanitization: No programmatic sanitization is applied to the variables before they are used in shell commands, relying entirely on the agent's adherence to instructions.
- [EXTERNAL_DOWNLOADS]: The skill downloads the Laravel framework and the Filament admin panel using standard package managers (Composer). It also directs users to download Laravel Herd from its official website (herd.laravel.com), which is a well-known service for Laravel development.
Audit Metadata