prompt
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill follows a standard instructional format to define a persona and a task. There are no attempts to bypass safety filters, extract system prompts, or override core agent instructions.
- [DATA_EXFILTRATION]: No network calls, credential harvesting, or sensitive file access patterns were identified. The skill only processes user-provided information for the intended design task.
- [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or system-level operations.
- [REMOTE_CODE_EXECUTION]: No external scripts are downloaded or executed. The skill relies entirely on the agent's internal generative capabilities.
Audit Metadata