The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to monitor application state and manage the build process.
Evidence includes instructions to run tail -f storage/logs/laravel.log | grep --line-buffered -E 'ERROR|CRITICAL|Exception', php artisan queue:work, and npm run build.
While these are standard commands for Laravel and Vite development workflows, they involve direct interaction with the system shell based on the skill's operational logic.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests and processes untrusted data generated by the application being tested.
Ingestion points: The agent reads content from storage/logs/laravel.log, browser console logs, and rendered web page content via agent-browser (e.g., Filament/Livewire pages).
Boundary markers: The skill does not specify the use of delimiters or specific instructions to ignore potential commands embedded within the logs or UI content it analyzes.
Capability inventory: The skill can execute shell commands (artisan, tail, npm), perform browser automation, and read local files.
Sanitization: There is no evidence of sanitization, filtering, or escaping of the application output before it is processed by the agent's logic.

verify