The Agent Skills Directory

[COMMAND_EXECUTION]: The Python scripts (completeness_checker.py, relationship_checker.py, schema_checker.py) process command-line arguments and JSON-formatted input using the standard library argparse and json modules. The scripts perform logic based on these inputs but do not use dangerous functions like eval(), exec(), or subprocess to execute external data or commands.\n- [DATA_EXFILTRATION]: The skill does not perform any network operations. It only reads from local ontology files and processes input provided through the agent. No access to sensitive system files or credentials was found.\n- [SAFE]: A comprehensive analysis of the logic, metadata, and configuration files confirms that the skill is a legitimate data validation utility. No evidence of prompt injection, obfuscation, or persistence mechanisms was found.

ontology-validator