AI Safety Guardrails

Output filter va policy layer. Inson faktori skilli bilan birga qo'llaniladi.

O‘rnatish

npx 21skills add data-ai/data-ai-core-20/ai-safety-guardrails

0 o‘rnatish

Agentlar

---
name: ai-safety-guardrails
description: Output filter va policy layer. (Data & AI to'plami, #20/20)
---

# AI Safety Guardrails

## Maqsad

Output filter va policy layer.

## Qachon ishlatish kerak

- Data & AI vazifalarida agent yordamida ishlashda
- Inson faktori skilli bilan birga qaror qabul qilganda

## Workflow

1. Kontekstni aniqlang.
2. Skill bo'yicha checklistni bajaring.
3. Inson review qiling.
4. Natijani repoga commit qiling.

## 21-skills ramkasi

- **#1** — [Inson faktori](/skill/human-judgment)
- **#20** — bu skill (Data & AI to'plami)