Welcome to the central documentation repository for all integrations with Palo Alto Networks Prisma AIRS (AI Runtime Security).
Prisma AIRS is a comprehensive AI security platform designed to protect the entire AI application lifecycle. It secures AI and traditional applications, agents, models, and datasets against a wide range of threats.
This repository contains step-by-step setup guides for integrating Prisma AIRS with various third-party MLOps, AI Gateway, and automation platforms. These integrations allow you to embed security directly into your AI workflows, enforcing policies and scanning for threats in real-time.
Below is a list of documented integrations. Each guide provides detailed prerequisites and configuration steps.
| Partner | Category | Documentation Link |
|---|---|---|
| LiteLLM | LLM Gateway | ./LiteLLM/README.md |
| n8n | Workflow Automation | ./n8n/README.md |
| Portkey | AI Gateway & Observability | ./Portkey/README.md |
| TrueFoundry | AI Gateway | ./TrueFoundry/README.md |
- AI Runtime Security (AIRS): A platform that provides inline, real-time security for AI applications, protecting against threats like prompt injection, malicious code, and sensitive data leakage.
- Strata Cloud Manager: The central management interface for configuring Prisma AIRS, including creating security profiles and generating API keys.
- Security Profile: A named configuration within Strata Cloud Manager that defines the specific threat detection rules and policies to be applied during a scan.
- Guardrail: A security control or policy implemented within a partner platform (like an AI Gateway) that leverages the Prisma AIRS API to scan and validate requests and responses.