The Agent Skills Directory

[SAFE]: No malicious behavior or patterns identified. The skill's purpose is restricted to text editing and style transformation based on publicly documented patterns.
[EXTERNAL_DOWNLOADS]: The README documentation suggests installing the skill via git clone from a GitHub repository. This is a transparent and standard installation procedure intended for the user.
[COMMAND_EXECUTION]: The README contains shell commands for manual installation (mkdir, git clone, cp). These are documentation for the user's setup process and are not executed automatically by the skill or the agent.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied text to perform its humanization task. While it has access to file modification tools, there is no evidence of malicious instructions or intent.
Ingestion points: User-provided text passed to the /humanizer command or direct prompts as seen in SKILL.md.
Boundary markers: Absent; the skill does not explicitly use delimiters to separate user input from its own instructions.
Capability inventory: Access to Read, Write, Edit, Grep, Glob, and AskUserQuestion tools.
Sanitization: None; the skill relies on the agent's default safety protocols when processing external content.

humanizer