skills/flora131/atomic/tool-design/Gen Agent Trust Hub

tool-design

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides educational content on designing deterministic contracts between agents and systems through optimized tool descriptions.
  • [COMMAND_EXECUTION]: The architectural reduction case study documents the use of a sandboxed bash tool as a design pattern, which is a conceptual example and not an implemented executable in the skill.
  • [EXTERNAL_DOWNLOADS]: The documentation mentions Vercel and the Model Context Protocol as examples of industry best practices; these are trusted sources and are used for reference only.
  • [SAFE]: The included Python script is a benign documentation utility that processes tool metadata using standard library modules for string and regex operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 03:51 AM