design-taste-frontend
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze untrusted data from user briefs, external URLs, and screenshots to inform design direction and code generation. This surface allows for indirect prompt injection where instructions embedded in external content could maliciously influence the agent. • Ingestion points: User briefs, provided URLs, and pasted screenshots (SKILL.md Section 0.A). • Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are specified for these inputs. • Capability inventory: Generates React/Next.js code, performs package installations (NPM/NPX), and utilizes image generation tools (SKILL.md Section 3, 4.8, Appendix A). • Sanitization: No sanitization, validation, or escaping of input data is defined before processing.
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install third-party dependencies from external registries and CDNs. • Evidence: Appendix A and B list commands for installing packages from established organizations including Microsoft, IBM, GitHub, and Shopify. • Remote references: References loading scripts from Shopify's CDN (https://cdn.shopify.com/shopifycloud/polaris.js) and assets from Simple Icons.
Audit Metadata