opendesign-codegen
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to translate user-provided design requirements into code, which inherently involves processing untrusted input (Indirect Prompt Injection surface). This risk is thoroughly mitigated by the 'Hard Rule Redlines' in SKILL.md and the 'Direct Code Checklist' in references/checklist.md, which strictly constrain the agent's output to specific component libraries and forbid the use of unsafe or arbitrary HTML/styles. These boundary markers effectively limit the capability of any injected instructions to influence the generated application logic.
Audit Metadata