BranchyMcChatFace

This application is an AI-powered chat interface featuring conversation branching capabilities. It combines a React frontend with a FastAPI backend, utilizes Docker for containerization, and includes features such as feedback submission with screenshot annotation.

Key Features

• Tree-like conversation structure with branching capabilities
• Support for multiple Language Models (LLMs)
• Dynamic conversation actions (deepening and broadening)
• Feedback mechanism with screenshot capture and annotation
• Containerized architecture for consistent deployment

Architecture

Backend (Python/FastAPI)

• Handles API requests for chat interactions and feedback submission
• Supports multiple LLMs, including a mock LLM for testing
• Utilizes environment variables for secure API key management

Frontend (React)

• Implements an intuitive tree-like conversation UI
• Supports dynamic conversation actions
• Features a comprehensive feedback mechanism

Docker

• Containerizes frontend and backend services
• Uses docker-compose for orchestration

Prerequisites

• Docker and Docker Compose
• Node.js v14+ (for local development)
• Python 3.9+ (for local development)

Installation

1. Clone the repository:

   git clone https://github.com/iterabloom/BranchyMcChatFace.git
   cd BranchyMcChatFace
   

2. Create a .env file in the root directory:

   OPENAI_API_KEY=your_openai_api_key_here
   ANTHROPIC_API_KEY=your_anthropic_api_key_here
   

3. Build and run Docker containers:

   docker-compose up --build
   

4. Access the application at http://localhost:3000

Usage Guide

1. Start a new conversation or continue an existing one.
2. Use the "DEEPEN" button to explore a conversation branch further.
3. Use the "BROADEN" button to create alternative conversation paths.
4. Submit feedback using the feedback button for screenshot annotations.

Development Setup

Backend

cd backend
pip install -r requirements.txt
uvicorn app:app --reload

Frontend

cd frontend
npm install
npm start

API Documentation

API endpoints are documented using OpenAPI (Swagger). Access the documentation at http://localhost:8000/docs when running the backend.

License

This project is licensed under the Mozilla Public License - see the LICENSE file for details.

Acknowledgments

• FastAPI and React communities for their excellent frameworks