skills/lev-os/agents/skill-creator/Gen Agent Trust Hub

skill-creator

Pass

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Python's subprocess module to run claude CLI commands and other internal scripts (scripts/run_eval.py, scripts/improve_description.py). This is the intended primary functionality for benchmarking and evaluating AI model performance against a set of prompts.
  • [EXTERNAL_DOWNLOADS]: The eval-viewer component includes a remote script reference to cdn.sheetjs.com (XLSX parsing library) in its HTML template. This is a well-known service used for the legitimate purpose of rendering spreadsheet data within the browser-based evaluation viewer.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 17, 2026, 01:24 AM