design-template-playful-geometric
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill's instructions strictly define a persona for a frontend engineer and designer. It does not attempt to override AI safety guidelines, bypass constraints, or extract system prompts. Findings: None.
- [DATA_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or unauthorized network operations were detected. The design tokens provided are standard hex codes and CSS values. Findings: None.
- [REMOTE_CODE_EXECUTION]: There are no commands that download or execute external scripts, and no dependencies on untrusted third-party packages. Findings: None.
- [OBFUSCATION]: The file content is entirely in plain-text markdown and YAML. No Base64, zero-width characters, or other encoding techniques were found. Findings: None.
- [DYNAMIC_CONTEXT_INJECTION]: The skill does not use shell command interpolation (
!command) or dynamic context placeholders that could lead to unauthorized command execution at load time. Findings: None. - [METADATA_POISONING]: The metadata accurately describes the skill's purpose and functionality. Findings: None.
- [INDIRECT_PROMPT_INJECTION]: While the skill is designed to interact with a user's codebase (via Read/Write/Edit), it does not process untrusted external data in a way that creates an injection surface beyond the inherent risks of a coding assistant. Findings: None.
Audit Metadata