react-vite-dashboard
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from Stitch screens and a DESIGN.md file to drive code generation and tool execution. While this is the core function of the skill, it creates a surface for indirect prompt injection if the design sources are maliciously crafted.
- Ingestion points: Stitch screen data via MCP and DESIGN.md tokens (SKILL.md workflow steps 2 and 4).
- Boundary markers: No explicit markers are used to delimit external design content from the agent's instructions.
- Capability inventory: Write for component generation and Bash for running linting tools.
- Sanitization: Absent; the skill assumes design inputs are trusted.
- [EXTERNAL_DOWNLOADS]: The skill fetches design assets and screen metadata from stitch.withgoogle.com. This is a well-known service associated with design tools.
- [COMMAND_EXECUTION]: The skill executes build and linting commands using npx, including the @google/design.md tool from a trusted organization.
Audit Metadata