How to Use
On page load, this mod loops through Zendesk conversations and uses ChatGPT to detect if any of the emails contain harmful or toxic content. If detected, the sidebar expands indicating the nature of the issue. The user can notify their manager via Slack, and also report a false positive so you can refine the prompt as needed.
Activate the Mod
Activate this mod. If you aren't logged into PixieBrix or need to create an account, you'll be prompted to do so before continuing the activation flow for this mod.
Our Built-in Open-AI configuration is set by default, so there's nothing to configure.
(Feel free to configure and add your own if you'd prefer, though.)
Click the purple Activate button.
Try it Out
Login to your Zendesk account and navigate to a ticket. On page load, you will see an alert that says "Checking for trust & safety..." PixieBrix is looping through the emails and sending the text as an input to the ChatGPT API. If a trust & safety issues is found in any of the emails, the sidebar will expand.
This is just the beginning! Once you activate your mod, the possibilities are endless. With PixieBrix, you can customize any mod to meet your needs or even build something entirely new. If you run into any trouble, need friendly support, or want to share your creations with the world, join our PixieBrix Slack Community.