laravel-audit
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes dynamic context injection (
!command) to execute diagnostic shell commands (e.g.,git rev-parse,grep composer.lock) at load time. These are used to identify the environment and framework version without model intervention. - [COMMAND_EXECUTION]: The skill requests access to powerful CLI tools such as
sshandrsyncto audit remote environments. To mitigate risks, it implements aPreToolUsehook (guard-readonly.sh) that uses a blacklist approach to deny destructive commands (e.g.,rm,mv,artisan migrate,composer install). - [DATA_EXFILTRATION]: By allowing
sshandrsync, the skill inherently possesses the capability to transfer data between hosts. This is the intended mechanism for auditing remote sources but represents a sensitive capability. - [PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks as it reads and analyzes external source code using tools like
catandgrep. Malicious instructions hidden within the audited codebase could potentially attempt to influence the agent's analysis or reporting behavior.
Audit Metadata