Detect prompt injection attacks in text
One step with Microsoft Azure. We can help you get access.
Look something up
Detect prompt injection attacks in text
with Microsoft Azure
Use Microsoft Azure: its Microsoft Azure ContentSafety Service Text:shield Prompt API lets you detect prompt injection attacks in text in a single step (POST /text:shieldPrompt). capabilities.work can help you get signed up and connected.
- How can I tell if a user prompt is trying to jailbreak my LLM app?
- Can Prompt Shields also scan grounding documents for indirect injection attacks?
Say it to your AI assistant
- Check user prompt {prompt} for injection attacks.
- Scan prompt {prompt} and documents {documents} for jailbreak or indirect prompt attacks.
For your developer
POST /text:shieldPrompt in Microsoft Azure ContentSafety Service Text:shield Prompt API. Full spec on APIs.io.