@349/starchild-design-pack

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified during the analysis. The skill operates as an informational resource using local scripts and data.
  • [PROMPT_INJECTION]: The skill instructions define a design-focused persona for the agent and provide styling defaults. No evidence of behavior overrides, safety filter bypasses, or 'jailbreak' patterns was found in the markdown files or CSV data.
  • [DATA_EXFILTRATION]: There is no access to sensitive system paths (e.g., SSH keys, cloud credentials) or network exfiltration logic. The documentation references a canonical guidelines source on a trusted Vercel GitHub repository.
  • [REMOTE_CODE_EXECUTION]: The Python scripts (core.py, design_system.py, search.py) use only standard library imports and do not download or execute remote code. There is no usage of unsafe primitives like eval() or exec() on untrusted input.
  • [COMMAND_EXECUTION]: No usage of subprocesses or system shell calls. The skill's file-writing capability is limited to generating design documentation in markdown format within a local project directory.
  • [DATA_EXPOSURE]: No hardcoded credentials or secrets were found in the dataset. CSV files contain design tokens, hex codes, and links to public documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 03:11 AM
Security Audit — agent-trust-hub — @349/starchild-design-pack