The Agent Skills Directory

[SAFE]: The skill functions as a static instruction set for converting JSON models into TUI layouts. It contains no commands that modify the system, access sensitive files, or execute remote code.
[SAFE]: Documentation references are restricted to the author's own GitHub repository (full-stack-skills), which constitutes legitimate vendor resource mapping.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates an attack surface for indirect prompt injection because it accepts dynamic input strings (e.g., error messages and properties) which are then rendered into output blocks. This is a functional requirement for UI generation, and no malicious exploitation patterns were found.
Ingestion points: Input JSON model in SKILL.md (specifically state.error and props).
Boundary markers: None defined for the rendering process.
Capability inventory: The skill generates drawing instructions (PENCIL_BATCH_DESIGN) intended for consumption by external drawing tools.
Sanitization: No explicit sanitization or escaping logic is provided for the interpolated input values.

tui-install