The Agent Skills Directory

[SAFE]: The skill's functionality is implemented using Python 3 and its standard library (argparse, json, os, sys). It does not require or install any external dependencies, minimizing supply chain risks.
[SAFE]: All inputs provided to the validation scripts are parsed using standard JSON libraries or explicit command-line argument handling, preventing common injection vulnerabilities within the script logic.
[SAFE]: File system access is limited to read-only operations on local reference files within the skill's directory or a closely-coupled sibling skill directory. No file writing or modification is performed.
[SAFE]: No network operations (exfiltration, remote downloads, or API calls) are present in any of the scripts or the skill's instructions.
[SAFE]: The skill uses explicit tool definitions (Read, Bash) as per platform requirements to execute its internal scripts, with no evidence of privilege escalation or persistence mechanisms.

ontology-validator