skill-creator

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill provides a legitimate framework for AI agent skill development, including workflow management, subagent orchestration for testing, and result visualization.
  • [EXTERNAL_DOWNLOADS]: The evaluation viewer uses assets from trusted sources, specifically Google Fonts and the SheetJS library for spreadsheet rendering. These are standard resources for web-based data visualization.
  • [COMMAND_EXECUTION]: Python scripts utilize subprocess calls to run the opencode CLI and manage a local web server via lsof. This is necessary for the skill's stated purpose of executing and reviewing skill performance.
  • [SAFE]: While there is a mismatch between the provided author (ederheisler) and the metadata author (anthropic), the behavior of the skill remains consistent with its described development utility and contains no malicious code.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:27 AM