design-taste-frontend-v2-experimental
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references official packages and documentation from well-known and trusted technology organizations including Microsoft, Google, IBM, Shopify, Atlassian, GitHub, and Vercel. These are standard development dependencies documented for setup purposes.
- [COMMAND_EXECUTION]: The documentation includes standard setup instructions using package managers (npm, yarn, npx) for legitimate, industry-standard frontend libraries.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data. Ingestion points: User-provided design briefs, vibe words, and external URLs (SKILL.md, Section 0.A). Boundary markers: The instructions do not define explicit delimiters for user input, relying on conversational context. Capability inventory: The skill provides templates for frontend code generation and instructions for package installation; it does not include capabilities for arbitrary shell execution, sensitive file access, or network exfiltration. Sanitization: No explicit sanitization or filtering logic is defined.
- [PROMPT_INJECTION]: The instructions include strict formatting constraints for the agent's output, such as specific character bans and aesthetic rules, which are legitimate functional requirements for the skill's intended purpose.
Audit Metadata