openai-agents

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references official and well-known packages from the npm registry, including @openai/agents and zod, within its templates and utility scripts. All sources are official and trusted repositories.
  • [COMMAND_EXECUTION]: Includes a utility script scripts/check-versions.sh designed to help developers verify their local SDK versions using the standard npm view command.
  • [SAFE]: Provides proactive security guidance, including specific templates for implementing safety guardrails to prevent prompt injection and instructions on secure API key management using ephemeral session tokens.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:49 AM