Output Filtering (Agent Safety Pattern)

Evolution Log

• 2026-04-24: 来自 24-Chapter-18-Guardrails-Safety-Patterns 的认知：A core agent safety pattern focused on screening and regulating outputs generated by intelligent agents to avoid releasing harmful content or instructions.