agentic-data-science-competition

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill content is entirely instructional and focused on data science competition workflows. No malicious patterns, obfuscation, or persistence mechanisms were identified. The author's use of standard automation tools (cronjobs, CLI) is consistent with the skill's stated purpose.
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download configuration and informational assets from trusted services such as GitHub and official Kaggle repositories. These downloads are transparent and target well-known, reputable technology platforms.
  • [COMMAND_EXECUTION]: Instructions guide the agent through the use of standard command-line tools (e.g., Kaggle CLI, Jupytext) for legitimate competition tasks. The skill includes best practices for error handling and validation to prevent unintended execution failures.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 03:52 PM