design-system-ui
Pass
Audited by Gen Agent Trust Hub on May 7, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides structured instructions for maintaining UI consistency by inspecting local project configuration files and existing components. No malicious code, obfuscation, or exfiltration patterns were found.
- [SAFE]: The reference material in references/design-moves.md contains standard design principles and does not include any executable code or dangerous instructions.
- [SAFE]: The skill uses the disable-model-invocation configuration, which restricts the agent from calling other tools, further reducing the potential attack surface.
- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted data from local project files. Ingestion points: nearby pages, components, tokens, and configuration files specified in SKILL.md. Boundary markers: Absent. Capability inventory: Writing frontend code; no network, filesystem-write (outside project files), or privileged system operations detected. Sanitization: Absent.
Audit Metadata