Go to file
Conor a9e0e9d21f
2024-01-08 11:31:03 +00:00
prompts Test adding a demo video 2024-01-07 18:34:21 +00:00
site Test adding a demo video 2024-01-07 18:34:21 +00:00
.gitignore Update site to work 2024-01-07 17:56:38 +00:00
LICENSE Update LICENSE 2024-01-08 11:29:27 +00:00
README.md Update README.md 2024-01-07 18:35:12 +00:00
chat.go Update site to work 2024-01-07 17:56:38 +00:00
go.mod Initial commit 2024-01-07 17:55:43 +00:00
go.sum Initial commit 2024-01-07 17:55:43 +00:00
main.go Initial commit 2024-01-07 17:55:43 +00:00
session.go Initial commit 2024-01-07 17:55:43 +00:00


Anti Jailbreak Site Chatbot with GPT-4 Turbo

This project is an attempt at preventing a "$1 chevy sale incident", by using two instances of ChatGPT in order to answer questions. It is very early stage and is overzealous at blocking innocent questions and inputs.



How to use

go run .

will launch a webserver on port 7129 or PORT if the environment variable is set. You need to provide your OPENAI_API_KEY as an environment variable or put it into a .env file. The chat interface is not particularly great and shows no indication that a reply is in the works, just wait a few seconds for it to appear.

What can I ask it?

It's currently trained on ConfigDN and so questions should be related to it to avoid rejection. TL:DR; ConfigDN is an Open Source configuration management and feature flag system with Go and JavaScript support currently. Ask questions about that. For more ideas see the prompts in the /prompts folder.

How it works

It first sends the users input to a GPT-4 Turbo instance, and this instance checks for any issues and rewrites the prompt safely for the second instance. The second instance then answers this question if it was considered "safe". There are also tripwires in the prompt to detect an attempted prompt leak.