space-wechat-layout
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill performs article formatting and local HTML generation without any suspicious behavior. It does not attempt to access sensitive files, environment variables, or external networks.
- [COMMAND_EXECUTION]: The instructions suggest using project-specific build or dev server commands (like Vite) to verify the generated HTML layout. This is standard developer workflow and poses no risk in this context.
- [PROMPT_INJECTION]: While the skill ingests untrusted user articles (a surface for indirect prompt injection), the instructions restrict the agent to layout-oriented optimization, preserving the original content and meaning.
- Ingestion points: User-provided article text collected via the prompt.
- Boundary markers: No specific boundary markers or delimiters are defined for the article content interpolation.
- Capability inventory: File system write (index.html), local command execution for build/server verification.
- Sanitization: The skill relies on natural language instructions to prevent substantial modification, rather than technical sanitization or escaping of the article content.
Audit Metadata