openai-agents-python

Pass

Audited by Gen Agent Trust Hub on May 4, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains instructional content and architectural patterns for using the OpenAI Agents SDK. It does not include executable code, external downloads, or commands that interact with the system in an unsafe manner.
  • [DATA_EXPOSURE]: No sensitive data exposure or exfiltration patterns were detected. The skill specifically recommends implementing guardrails to prevent external data transmission as a safety best practice.
  • [PROMPT_INJECTION]: No prompt injection, role-play bypass, or behavior override patterns were found in the instructions or metadata.
  • [EXTERNAL_DOWNLOADS]: The skill references an official GitHub repository from a well-known and trusted technology organization (OpenAI).
Audit Metadata
Risk Level
SAFE
Analyzed
May 4, 2026, 12:15 PM
Security Audit — agent-trust-hub — openai-agents-python