web-design-guidelines
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill fetches instruction content from 'https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md'. Because 'vercel-labs' is a trusted organization, the download itself is low risk, but the dependency remains unverifiable at runtime.
- [PROMPT_INJECTION] (HIGH): The skill exhibits a high-risk surface for Indirect Prompt Injection. Ingestion points: External instruction file (command.md) and untrusted local files (UI code). Boundary markers: Absent; the skill lacks delimiters or instructions to ignore embedded commands in audited files. Capability inventory: File system read access and network access via WebFetch. Sanitization: None. Risk: A compromised remote source or an audited file could inject instructions to exfiltrate sensitive data from the local environment using the agent's network capabilities.
Recommendations
- AI detected serious security threats
Audit Metadata