tui-cell
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides structured guidance and examples for rendering a UI component in ASCII and generating corresponding design specifications for use with Pencil MCP. No malicious code, unauthorized network access, or credential theft patterns were detected.
- [NO_CODE]: The skill consists entirely of markdown instructions and JSON/text examples. No external scripts or executable files are included or referenced for execution.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to process untrusted data (the 'Input Model') to produce structured output blocks (COMPONENT_SPEC, PENCIL_BATCH_DESIGN).
- Ingestion points: Untrusted data enters the agent's context through the 'props', 'state', and 'layout' fields of the defined JSON Input Model.
- Boundary markers: The instructions do not define specific delimiters or instructions for the agent to disregard potential commands embedded within the input data.
- Capability inventory: The skill generates specifications and batch design operations intended to be consumed by the Pencil MCP tool.
- Sanitization: No input validation or output escaping mechanisms are described to prevent malicious content in the input from appearing in the generated specifications.
Audit Metadata