The Agent Skills Directory

[SAFE]: The skill defines a benign process for visual design experimentation and documentation. No indicators of prompt injection, data exfiltration, obfuscation, or unauthorized command execution were found.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided design questions to influence code generation and documentation updates.
Ingestion points: User design questions provided in chat.
Boundary markers: None explicitly defined in the instructions.
Capability inventory: The agent is instructed to implement code variants and edit project documentation files (DIRECTION.md, TOKENS.md).
Sanitization: No specific sanitization or validation logic is defined for the user input.
Analysis: This interaction represents the primary intended functionality of the skill. While it defines an attack surface common to AI agents, the scope is limited to UI prototyping and design documentation, which is considered low risk in this context.

ui-prototype