The Agent Skills Directory

[SAFE]: The skill is entirely instructional and focused on UI/UX quality control. There is no evidence of code execution, network exfiltration, or persistence mechanisms.
[EXTERNAL_DOWNLOADS]: The skill references external design resources such as 'land-book.com', 'curated.design', and 'componentry.fun'. These are utilized as informational references for the agent to guide its visual output and do not involve executable code or automated package installation.
[PROMPT_INJECTION]: The instructions include firm constraints (e.g., 'Non-Negotiable', 'Hard bans') to maintain output quality, but these do not attempt to bypass safety guardrails or subvert agent instructions outside the scope of frontend design.

uncodex