frontend-design
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface by ingesting untrusted project configuration data.\n
- Ingestion points: Reads
.better-web-ui.md,.better-ui.md, and.impeccable.mdfrom the project root (SKILL.md).\n - Boundary markers: Absent; no delimiters or 'ignore embedded instructions' warnings are applied to the gathered context.\n
- Capability inventory: The skill is capable of executing shell commands for package management and modifying the application codebase.\n
- Sanitization: Absent; the content is utilized directly to shape the agent's output.\n- [PROMPT_INJECTION]: Static analysis identified potential action concealment in
reference/marketing-copywriting.md. Review confirms this is a false positive; the terms 'concealment' and 'hide' appear solely within guidelines that discourage 'dark patterns' in UI design, advocating for interface honesty rather than attempting to override agent behavior.
Audit Metadata