php-framework-audit
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's core purpose is to perform white-box security audits on external PHP source code. This creates a surface for indirect prompt injection attacks, where a malicious project being audited could contain instructions hidden in comments or configuration files designed to subvert the agent's logic.
- Ingestion points: The skill processes PHP source files, framework configuration files (
composer.json,.env,wp-config.php), and directory structures across Laravel, ThinkPHP, WordPress, Symfony, Yii2, and CodeIgniter projects. - Boundary markers: The instructions lack explicit boundary markers or instructions to treat data as untrusted, which could lead the agent to follow instructions embedded within the code being analyzed.
- Capability inventory: The skill instructions focus on code analysis and pattern matching using system tools (e.g., grep, find), requiring read access to the entire project directory.
- Sanitization: No sanitization or validation mechanisms are described for the content being processed during the audit.
- [NO_CODE]: The skill consists entirely of Markdown-based documentation and audit checklists. It does not include any scripts, executables, or automated commands that would execute during the audit process, reducing the risk of direct malicious behavior.
Audit Metadata