# GCP Agent Model Armor Shield

> Configures Model Armor security policies (Prompt Injection, Jailbreak, RAI filters).

- Skill: `googlecloudplatform/gcp-agent-model-armor-shield` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add googlecloudplatform/gcp-agent-model-armor-shield`
- Raw SKILL.md: https://api.skillmd.com/api/skills/googlecloudplatform/gcp-agent-model-armor-shield/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: GoogleCloudPlatform (https://skillmd.com/u/googlecloudplatform)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/googlecloudplatform/gcp-agent-model-armor-shield

---


# gcp-agent-model-armor-shield

This skill configures Model Armor as an intelligent firewall for your AI agents. Grounded in `security_blog.md`, it protects against prompt injection, jailbreaking, malicious URLs, and links to SDP templates for data privacy.

## Usage

Ask Antigravity to:
- "Configure Model Armor for my agent"
- "Add prompt injection protection to my security policy"
- "Set up RAI filters for hate speech and harassment"
- "Link my SDP templates to Model Armor via Terraform"

## Protection Layers

1. **Detection Filters**: Prompt Injection, Jailbreak, and Malicious URI detection.
2. **RAI Settings**: Configurable confidence levels for Hate Speech, Harassment, Sexually Explicit, and Dangerous Content.
3. **SDP Integration**: Hooks for Advanced Sensitive Data Protection (linking to Inspect/De-identify templates).

## Terraform Boilerplate

Refer to `resources/model_armor.tf` for the standard configuration.