The Agent Skills Directory

[SAFE]: The skill is purely instructional and provides guidance on thumbnail design strategy. It does not contain any executable code, shell commands, or network requests.
[DATA_EXPOSURE]: The skill checks for the existence of and reads .agents/youtube-context.md. This is a local project file used to provide relevant context for the AI agent's responses. There is no mechanism within the skill to exfiltrate this data or access sensitive system files like SSH keys or environment variables.
[PROMPT_INJECTION]: The instructions do not contain patterns aimed at bypassing AI safety filters or overriding system instructions. It uses a professional 'YouTube thumbnail strategist' persona that remains within the bounds of standard agent behavior.

thumbnail-design