skill-creator

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Python's subprocess module across multiple files (run_eval.py, improve_description.py, validate-skill.py) to execute the claude CLI and Python scripts. This behavior is documented and necessary for its core features, such as benchmarking skill triggering and performing rubric-based validation.
  • [EXTERNAL_DOWNLOADS]: The evaluation viewer component (viewer.html) includes a reference to the SheetJS library hosted on cdn.sheetjs.com. This is used to render spreadsheet files locally within the interactive report generated by the skill.
  • [DATA_EXPOSURE]: The generate_review.py script utilizes the Python HTTPServer module to host a local server on 127.0.0.1. This server facilitates the review of evaluation results by serving files from the user's specific workspace directory to a web browser.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 01:04 PM