vibe-coding-css-setup
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project documentation files to generate code.
- Ingestion points: The skill reads
./docs/DESIGN_SYSTEM.mdand./docs/TECH_STACK.mdto extract CSS tokens and framework preferences (SKILL.md). - Boundary markers: There are no explicit delimiters or instructions to the agent to treat the content of these files as potentially adversarial data.
- Capability inventory: The skill has the capability to write to critical project files including
globals.css,tailwind.config.js, and root layout components (app/layout.tsx). - Sanitization: The instructions direct the agent to copy blocks of text directly from the documentation into the generated CSS and JS files without validation or escaping.
- [EXTERNAL_DOWNLOADS]: The skill configures projects to fetch typography assets from Google Fonts via
@importdirectives in the generated CSS. This targets the well-knownfonts.googleapis.comdomain.
Audit Metadata