The Agent Skills Directory

[SAFE]: The core logic of the skill is dedicated to linguistic transformation and readability improvement. It explicitly instructs the agent to preserve technical precision and all meaning while only simplifying the phrasing and structure, which is a benign and useful functionality.
[EXTERNAL_DOWNLOADS]: The README provides installation instructions using the npx package runner to fetch the skill from the author's repository. This is a standard installation method for skills from this author.
Evidence: npx skills add shihyuho/skills --skill plain -g in README.md.
[PROMPT_INJECTION]: The skill processes untrusted input from users or files, which is a standard surface for indirect prompt injection. However, since the skill is designed to rephrase text into plain language and does not have access to dangerous system tools or network operations, the risk is minimal.
Ingestion points: SKILL.md identifies targets as pasted text, file paths, or conversation history.
Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions for the target content.
Capability inventory: The skill performs text manipulation only and does not invoke subprocesses, network tools, or sensitive file writes.
Sanitization: No explicit input sanitization instructions are present.

plain