CodeSynthAI is an experimental open-source project that explores the potential of AI-powered Python scripting. By combining Large Language Models (LLMs) with a multi-agent system.
- About the Project
- Screenshots
- Tech Stack
- Features
- Environment Variables
- Getting Started
- Prerequisites
- Installation
- Usage
- Roadmap
- Contributing
CodeSynthAI is an experimental proof-of-concept (POC) project that explores the potential of AI-assisted code generation using a multi-agent system. It aims to streamline the process of creating Python scripts by leveraging the power of Large Language Models (LLMs) and collaborative AI agents.
Please note that CodeSynthAI is currently a work in progress (WIP) and is in its early stages of development. The project serves as an initial exploration of the concept and is not yet a fully-fledged solution. The code generated by CodeSynthAI may vary in quality and reliability, and it should be thoroughly reviewed and tested before being used in any production environment.
Language
Libraries and Tools
- AI-assisted Python script generation
- Multi-agent collaboration for code refinement
- Integration with OpenAI embeddings and Pinecone vector store
- User-friendly Streamlit interface
- Iterative development process
TBA
- Python 3.7+
- OpenAI API key
- Anthropic API key
- Pinecone API key
- Clone the repository:
git clone https://github.com/peterzervas/CodeSynthAI.git cd CodeSynthAI - Clone the repository:
pip install -r requirements.txt
- Set up the necessary API keys:
OpenAI: Set the OPENAI_API_KEY environment variable. Anthropic: Set the ANTHROPIC_API_KEY environment variable. Pinecone: Set the PINECONE_API_KEY and PINECONE_ENVIRONMENT environment variables.
- Start the Streamlit app
streamlit run app.py or python -m streamlit run app.py
Enter your script requirements in the text area provided in the Streamlit app. Click on "Generate Code" to initiate the code generation process. Review and confirm the refined requirements. Click on "Start Scripting" to begin the iterative scripting process. Provide feedback on the generated code by selecting "👍 Thumbs Up" or "👎 Thumbs Down" and click "Submit Feedback and Add to Vector Store" to store the code snippet and feedback.
- Initial proof-of-concept release
- Requirements
- Expand support to other programming languages
- Improve code quality and reliability
- Integrate with version control systems
- Explore more advanced AI techniques and models
Made with contrib.rocks.
Contributions are always welcome!
See contributing.md for ways to get started.
Distributed under the MIT License. See LICENSE for more information.