The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references the @rephrasy/mcp package, which is a Node.js package provided by the vendor for structural humanization of AI text.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied drafts, creating an ingestion point for untrusted data that could contain malicious instructions.
Ingestion points: User-provided text drafts and writing samples intended for humanization.
Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore embedded prompts within the processed text.
Capability inventory: The skill uses text transformation logic and makes external tool calls to the humanize function in the @rephrasy/mcp package.
Sanitization: No documented sanitization or input filtering is performed on the user text before humanization.

bypass-turnitin