frontend-design
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from external design galleries which could potentially contain malicious instructions.\n
- Ingestion points: A research subagent is instructed to access and analyze content from external websites, including
threeui.com,awwwards.com, andgodly.website, using scraping and vision capabilities (SKILL.md).\n - Boundary markers: There are no instructions provided to separate the external data from the agent's internal instructions or to warn the agent about ignoring embedded prompts in the retrieved content.\n
- Capability inventory: The skill has the ability to generate production-ready code (HTML, CSS, JavaScript, React, Vue) and uses the
invoke_subagenttool to perform complex tasks.\n - Sanitization: The skill lacks any steps for validating, filtering, or sanitizing the content extracted from the design websites before it is used to guide the final code implementation.
Audit Metadata