The Agent Skills Directory

[SAFE]: The skill is a purely instructional text-processing tool. It focuses on improving writing style and rhythm without performing suspicious network operations, hardcoding credentials, or attempting to hide its logic.
[PROMPT_INJECTION]: The skill ingests untrusted input via the $ARGUMENTS variable and the --file parameter, interpolating this data directly into the agent's context. This creates a potential surface for indirect prompt injection.
Ingestion points: User-supplied text in arguments and file content read through the --file flag in SKILL.md.
Boundary markers: Absent; the skill lacks delimiters or instructions to ignore embedded commands within the text being humanized.
Capability inventory: The skill utilizes Read, Write, Edit, Grep, and Glob tools, granting the agent the ability to modify local files based on instructions.
Sanitization: None; the input is processed directly without escaping or validation logic.
[NO_CODE]: The skill is implemented as a set of sophisticated prompt instructions within a Markdown structure. It does not include any Python, JavaScript, or shell script files, minimizing the risk of traditional code-based exploits.

humanizer