What it does
It caches the resulting verdict and can block high-scoring requests in enforce mode; provider errors fail open.
Integrations
jev-edge
jev-edge puts semantic prompt-injection and abuse checks in front of traffic handled by nginx, OpenResty and several gateways.
Bookmark: jev-edge
Keep this in your collection.
Leave a noteWhat would you try with this?
: jev-edge
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Before touching your server, write down what your application does and describe typical user messages. Your developer adds this description to your server settings. They also supply an access key that connects your web server to the external check service.
Ask your developer to start in monitor mode first. This mode records safety scores in your logs without stopping real visitors. Once the scores look trustworthy, your developer can tell the server to block harmful text. Automated checks can still make mistakes, and traffic continues through if the checking service goes offline.