review-frontend
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains no instructions designed to override safety filters, bypass constraints, or extract system prompts. The instructions are focused entirely on code review criteria.
- [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or mechanisms for data harvesting were detected.
- [REMOTE_CODE_EXECUTION]: There are no patterns involving remote script downloads (e.g., curl|bash) or dynamic execution of untrusted code. Mentioned libraries like react-window and lodash-es are provided as examples for the reviewer, not as runtime dependencies.
- [EXTERNAL_DOWNLOADS]: No external URLs or remote resources are referenced for download or execution.
- [COMMAND_EXECUTION]: The skill does not invoke shell commands, subprocesses, or system-level APIs.
- [OBFUSCATION]: The content is written in clear, plain markdown with no encoded strings, hidden characters, or steganographic patterns.
Audit Metadata