The Agent Skills Directory

[SAFE]: Analysis of the 'human-ai' skill components and instructions confirms the skill is dedicated to text stylistic transformation and does not contain malicious code.
[COMMAND_EXECUTION]: The included Python script scripts/measure.py is used to analyze text metrics such as TTR and entropy. It uses standard libraries and contains no logic for network access, file system writes, or subprocess execution.
[DATA_EXFILTRATION]: There are no indicators of data exfiltration. The skill does not access sensitive local files or transmit information to external endpoints.
[PROMPT_INJECTION]: The instructions for rewriting text are stylistic in nature and do not attempt to override the AI agent's core instructions or safety guidelines.
[EXTERNAL_DOWNLOADS]: All external URLs mentioned in the skill are informational links to documentation, research papers, and relevant community repositories.

human-ai