https://github.com/mtib/ai-chat
Frontend for chatting with OpenAI
https://github.com/mtib/ai-chat
Last synced: 4 months ago
JSON representation
Frontend for chatting with OpenAI
- Host: GitHub
- URL: https://github.com/mtib/ai-chat
- Owner: mtib
- Created: 2025-04-05T08:59:26.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-04-18T23:11:39.000Z (about 1 year ago)
- Last Synced: 2025-04-19T02:28:36.879Z (about 1 year ago)
- Language: TypeScript
- Homepage: http://chat.mtib.dev/
- Size: 620 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# ai-chat
A chat application with support for OpenAI models and custom server assistants.
## Features
- Chat with OpenAI's latest models
- Generate images with DALL-E
- Support for custom server assistants with contextual data
- Message starring for important information
- Export and import conversations
- Conversation search and management
- PWA support for offline access
## Server Assistant API

You can add your own assistant using the server assistant API. This allows you to:
1. Connect to your own assistant server with custom knowledge
2. Use your server's system prompt for conversations
3. Enhance conversations with contextual data from your server
### Setup
To connect a server assistant:
1. Create a new conversation and select "Server Assistant"
2. Enter your server URL and authentication token
3. The system prompt will be loaded from your server
### How it works
When you chat with a server assistant:
- The app fetches the system prompt from your server's `/config` endpoint
- Your messages are embedded and sent to the server's `/search` endpoint
- Relevant context from your server is added to OpenAI requests
- Your conversation history remains private and doesn't include the assistant's context
## Additional Documentation
- [Assistant Server Implementation](assistant_server/README.md)
- [Assistant Server Setup Guide](assistant_server/SETUP.md)
- [JSON Server with Redis](json_server/README.md)