prototyping-ui-directions
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches design references by cloning Git repositories from platforms like GitHub and capturing screenshots of established websites (e.g., Linear, Stripe, Vercel) during the research phase. All external acquisitions are explicitly gated by user confirmation.
- [COMMAND_EXECUTION]: Utilizes system commands like git clone for asset acquisition. The generated prototype packages are designed for local research and may include instructions for users to run standard package managers like npm to view the resulting mocks.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external code and documentation from third-party sources to extract UI patterns. This ingestion of untrusted data into the agent's context represents a potential attack surface for indirect prompt injection, although the use of structured extraction templates and standard agent guardrails reduces the severity of this risk.
Audit Metadata