development
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The skill instructs the agent to run standard development lifecycle commands such as formatters, linters, type checkers, and tests. These are routine operations for code validation.
- [EXTERNAL_DOWNLOADS] (SAFE): The skill references established package managers (uv, pnpm) and reputable libraries (Pydantic, React Query, Drizzle). No downloads from untrusted or malicious sources are present.
- [PROMPT_INJECTION] (SAFE): No instructions designed to override safety filters or bypass constraints were detected. The skill uses professional engineering personas and standard workflow principles.
- [DATA_EXFILTRATION] (SAFE): There are no patterns indicating the exfiltration of secrets or sensitive files. The use of web_search and imagegen is aligned with the described web design tasks.
Audit Metadata