The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses a set of instructions to guide the agent in generating design systems. No patterns of behavioral override, safety bypass, or role-play jailbreaks were detected.
[DATA_EXFILTRATION]: The Python scripts (core.py, search.py, design_system.py) only perform local file read/write operations within the skill's directory and a project-specific design-system/ folder. No network operations (such as requests or curl) are present in the code.
[COMMAND_EXECUTION]: The skill documents its own usage via a CLI tool (search.py). All command examples are restricted to local script execution for the tool's intended functionality. There are no instances of arbitrary command injection or shell piping of remote content.
[REMOTE_CODE_EXECUTION]: The Python logic uses standard libraries to process CSV data and generate Markdown files. It does not use eval(), exec(), or any mechanism to download and run external scripts.
[EXTERNAL_DOWNLOADS]: The skill references documentation for popular frameworks (Next.js, Astro, Tailwind, etc.) and Google Fonts URLs. These are informative and point to well-known, trusted technology services. No runtime downloads or package installations are executed by the skill's code.
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or sensitive credentials were found in the scripts or data files.
[OBFUSCATION]: The code and data are written in clear, readable text. No Base64, hex-encoding, or hidden Unicode characters were identified.

ui-ux-pro-max