The Agent Skills Directory

[SAFE]: The skill operates entirely within its declared scope of content editing. No evidence of prompt injection, obfuscation, or malicious intent was found in the instructions or reference files. The skill focuses on linguistic analysis and voice transformation without unauthorized system interaction.
[DYNAMIC_EXECUTION]: The skill includes a local Python script (scripts/humanizer_scorer.py) used to audit text for AI patterns. The script utilizes only standard Python libraries (sys, re, json, math, collections, argparse) and performs regex analysis on text provided by the user. It does not spawn subprocesses, compile code at runtime, or interact with external APIs.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text drafts, which is an inherent surface for indirect prompt injection. However, the skill lacks dangerous capabilities such as network access, file writing, or system command execution, which significantly mitigates the risk.
Ingestion points: User-provided text content drafts and brand voice notes in SKILL.md.
Boundary markers: None explicitly defined.
Capability inventory: Text transformation and local script analysis with no side effects.
Sanitization: None present, as the primary task is to rewrite the ingested text.

content-humanizer