What it does
The checks use typed yes-or-no and severity questions.
Integrations
agentgateway Jev guardrail example
An agentgateway example that screens LLM requests and responses with Jev guardrails.
Bookmark: agentgateway Jev guardrail example
Keep this in your collection.
Leave a noteWhat would you try with this?
: agentgateway Jev guardrail example
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Start by listing the trick questions, private secrets, or harmful topics you want to block. You can decide how strict the score limits should be. A developer can then set up this example between your users and an AI model like gpt-5.6-luna.
Your developer will need account access codes from OpenAI and TypeSafe to link the tools together. Remember that user questions and AI answers travel out to TypeSafe so the external service can review them. If the safety checker runs into a glitch, it blocks the request.