---
title: "AI Policies"
description: "Kyverno policies for governing AI agents and workflows. Includes prompt safety rules, AI skill authorization policies, and the AI Policy Pack for LLM-integrated workloads."
diataxis: explanation
applies_to:
  product: "nirmata-ai-agents"
audience: ["platform-engineer"]
last_updated: 2026-03-25
url: https://docs.nirmata.io/docs/policy-sets/ai-policies/
---


Kyverno policies for governing AI agents, MCP servers, and LLM-integrated workloads.

## What's Covered

- **Prompt guardrails** — Enforce safe prompt patterns and block prohibited inputs
- **Skill authorization** — Control which nctl AI skills can be invoked
- **AI Policy Pack** — Bundled policies for LLM workload governance
- **MCP server policies** — Access control and audit rules for MCP-compatible gateways

All AI policies are available in the [Nirmata policy library on GitHub](https://github.com/nirmata/kyverno-policies/tree/main/ai).