The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes untrusted text provided by users or retrieved from local files without explicit instruction isolation.
Ingestion points: Untrusted text from user input or files accessed via Read and Glob tools.
Boundary markers: Absent. The prompt does not specify delimiters (e.g., XML tags or block markers) to separate the data being humanized from its internal instructions.
Capability inventory: The skill is permitted to use Read, Write, Edit, Grep, and Glob tools, which could potentially be misused to modify files if a malicious instruction embedded in the input text were followed.
Sanitization: No input validation or filtering of instructions is implemented.
[NO_CODE]: The skill is composed entirely of markdown-based instructions and metadata; it does not ship with any executable scripts, binaries, or external dependencies.
[SAFE]: No malicious behavior, data exfiltration, or unauthorized network operations were detected. External references point to well-known, reputable sources such as Wikipedia.

humanizer