Guardrails & Safety card artwork

Guardrails & Safety

Epic

by OpenAI

Implement safety guardrails with input/output filtering, jailbreak detection, and policy enforcement.

9,340downloads
Security Verified
Security
Sign In to Download View on GitHub

About This Skill

Build robust safety systems for AI applications. Includes input sanitization, output filtering, jailbreak attempt detection, PII redaction, topic restriction enforcement, and comprehensive audit logging. Configurable policy engine for custom safety rules.

Compatibility

GPT-4ClaudeGemini