The Agent Skills Directory

[SAFE]: The skill operates as a complex text-processing pipeline. It restricts file operations to a dedicated _workspace directory and uses the standard Agent tool for coordination.
[COMMAND_EXECUTION]: The orchestrator uses the Agent tool to delegate tasks and the Glob tool for file discovery. It explicitly prohibits the use of potentially unsafe shell commands like ls for path resolution.
[EXTERNAL_DOWNLOADS]: The documentation cites academic papers from arXiv and refers to well-known technology companies (Google, DeepL, Lionbridge). These are informative references and do not involve the download or execution of untrusted remote code.
[NO_CODE]: The Python scripts provided for metric calculation (metrics.py and metrics_v2.py) are implemented using standard libraries (re, json, math, argparse). They perform static analysis of the input text using regular expressions and do not contain any dangerous system calls, network access, or dynamic code execution.

humanize-korean