prototype
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to provide instructions for running generated prototypes using system commands such as pnpm, python, and bun in the project's task runner environment.- [INDIRECT_PROMPT_INJECTION]: The workflow generates executable prototype code based on user-supplied questions, which could be exploited by an attacker to cause the agent to generate malicious code.
- Ingestion points: User-provided design questions used to determine prototype logic and UI structure.
- Boundary markers: No explicit delimiters or instructions are specified to help the agent distinguish between the design question and potential embedded malicious instructions.
- Capability inventory: The skill allows for writing prototype files (HTML, TSX) and suggests shell commands for execution.
- Sanitization: No sanitization or validation logic is defined for the input question before it influences code generation.
Audit Metadata