The Agent Skills Directory

[SAFE]: The skill demonstrates secure coding practices, specifically using yaml.safe_load() in its validation scripts to prevent potential YAML deserialization vulnerabilities.\n- [SAFE]: No hardcoded credentials, API keys, or attempts to access sensitive file paths (such as SSH keys or environment variables) were found within the skill or its associated references.\n- [SAFE]: The included Python scripts (scripts/validate_ontology.py and scripts/blocking_report.py) use only standard libraries and the well-known pyyaml package, with no network activity or remote code execution detected.\n- [SAFE]: The workflow design prioritizes deterministic code for identity merging and data validation, effectively mitigating potential indirect prompt injection risks from unstructured source text.

kg-builder