The Agent Skills Directory

[SAFE]: No security issues were identified in the skill instructions or logic. The operations are limited to reading local theme files and applying visual styling to user documents.
[DATA_EXFILTRATION]: There are no network operations, external requests, or mechanisms present that could lead to data exfiltration. The skill maintains data locality.
[PROMPT_INJECTION]: The skill processes user-provided inputs and artifacts for theme generation and application. While this constitutes an indirect prompt injection surface, the risk is negligible as the agent's actions are strictly constrained to visual styling.
[COMMAND_EXECUTION]: No shell commands, system calls, or script execution patterns were detected in the provided skill files.

theme-factory