The Agent Skills Directory

[SAFE]: The skill promotes standard, well-documented security best practices for the Laravel framework, including the use of policies for authorization and the proper handling of environment variables.
[EXTERNAL_DOWNLOADS]: The skill mentions the spatie/laravel-permission package as a recommended option for role-based access control. This is a well-known and widely trusted library in the Laravel ecosystem.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process content from application policy files (app/Policies/**/*.php) and configuration files (config/*.php). This represents a standard attack surface for indirect prompt injection where the agent could potentially process instructions embedded in the codebase.
Ingestion points: PHP policy and configuration files targeted by the skill's triggers.
Boundary markers: None identified in the skill instructions to delimit untrusted code content.
Capability inventory: The skill suggests generating and modifying code and executing php artisan commands.
Sanitization: No specific sanitization or validation of the ingested file content is mentioned.

laravel-security