BuildShip Logo
BuildShip Templates

Model Armor

Screens LLM prompts and responses for prompt injection, jailbreaks, harmful content, sensitive data leaks, malicious URLs, and CSAM using Google Cloud Model Armor. Supports multi-language detection and works with any LLM platform.

0

Report this node

Select the reason for reporting

Describe the issue in detail

Share node

Link to node

https://templates.buildship.com/node/community_hE6gnkhjNq01/

Inputs

GCP Project ID

escriptwallet

Location

us-west1

Template ID

test-template

Mode

sanitizeUserPrompt

Text

Check out this link: http://malware.testing.google.test/testing/malware/

User Prompt Context

Enable Multi-Language Detection

True
False

Source Language Code

This is a static example using sample inputs. Copy the node and paste it into your BuildShip flow to run it with your own values.

Output