The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown-based design guidelines and instructional content. No executable scripts, binaries, or automated shell commands are present in any of the files.
[PROMPT_INJECTION]: The instructions do not attempt to bypass safety filters, override system instructions, or extract system prompts. The 'Conflict Resolution' section in SKILL.md is standard for governing the priority of different design rules and does not contain malicious override patterns.
[DATA_EXFILTRATION]: There is no evidence of network access tools (curl, wget, fetch) or sensitive local file system access (e.g., .ssh, .aws, .env).
[REMOTE_CODE_EXECUTION]: No package managers (npm, pip) or external dependency fetching mechanisms are used. The skill is entirely self-contained within its documentation files.
[OBFUSCATION]: Analysis of all files, including the detailed accessibility and typography guides, shows no hidden characters, Base64-encoded payloads, or homoglyph attacks.
[DYNAMIC_CONTEXT_INJECTION]: The skill does not use the !command syntax in SKILL.md for execution at load time.

typeui-fundamentals