prototype
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions suggest defining project-specific execution commands (e.g., pnpm, python, or bun) to facilitate running the generated prototype. This is standard developer workflow but involves defining new command execution entry points within the project environment.
- [INDIRECT_PROMPT_INJECTION]: The skill interprets user prompts and project context to generate executable code (HTML/JS or TSX), which represents a potential attack surface. However, the instructions mitigate this by directing the agent to use isolated logic modules, stub mutations, and avoid real database persistence.
Audit Metadata