The Agent Skills Directory

[SAFE]: The skill implements a set of technical guidelines and utility scripts (data_doctor.py, quality_gate.py, safe_split.py, check_access_level.py) designed to handle structured data processing. These scripts operate locally and do not perform unauthorized network operations or system modifications.
[SAFE]: A security-focused component, check_access_level.py, is included to enforce data governance. It prevents high-sensitivity data (PII/raw) from being used in public-facing outputs such as research papers or code repositories, demonstrating a security-first design.
[SAFE]: All external tools referenced in the documentation, including pandas, polars, duckdb, cleanlab, and evidently, are well-known and reputable services within the technology and data science communities.
[SAFE]: Recommendations for managing API credentials (e.g., Kaggle tokens) follow standard security practices, such as suggesting the use of specific file paths with restricted (600) permissions.
[SAFE]: No patterns of prompt injection, code obfuscation, or remote code execution from untrusted sources were found in any of the analyzed files.

light-data-engineering