contact-page-generator
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No evidence of instructions designed to bypass agent constraints or safety filters was found. The instructions are focused on providing structural guidance for web design.- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill checks for local project context files to maintain brand consistency, which is a standard and safe practice. It does not contain any hardcoded credentials or instructions to send data to external servers.- [REMOTE_CODE_EXECUTION]: No patterns for downloading or executing external scripts or packages were detected. The skill is entirely descriptive.- [OBFUSCATION]: All instructions and guidelines are provided in clear, human-readable text. No hidden or encoded content was found.- [INDIRECT_PROMPT_INJECTION]: While the skill reads project context files which could technically contain untrusted data, it lacks any high-privilege capabilities (like shell access or network tools) that could be exploited by such data. The surface is minimal and safe.
Audit Metadata