php-codeigniter-audit
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs static analysis on local project files specified by the user to identify security misconfigurations and coding flaws. No network operations or unauthorized system access patterns were detected.- [PROMPT_INJECTION]: The skill processes untrusted source code as its primary function, which creates a surface for indirect prompt injection. Attackers could potentially place malicious instructions within the source code to influence the auditor's findings. * Ingestion points: Reads all files within the provided project directory. * Boundary markers: None present; the agent processes raw code directly. * Capability inventory: File system read access for analysis and write access for report generation. * Sanitization: No explicit sanitization or instruction-filtering is applied to the analyzed code content.
Audit Metadata