design-taste-frontend
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill features an indirect prompt injection surface by requiring the agent to read and act upon the project's package.json file.
- Ingestion points: The agent parses the local package.json file to check for dependencies.
- Boundary markers: The skill does not specify markers or instructions to isolate the file content from the agent's primary instructions.
- Capability inventory: The agent is authorized to suggest installation commands and generate complex frontend code.
- Sanitization: There is no evidence of validation or filtering for the metadata retrieved from the package file.
- [EXTERNAL_DOWNLOADS]: The skill encourages the use of well-known, high-reputation libraries such as framer-motion, lucide-react, zustand, and Radix UI. It directs the agent to provide installation commands for these packages if they are not already present in the user's environment.
Audit Metadata