AutoCoder

AutoCoder is a cutting-edge programming assistant powered by GPT-4/PaLM. It operates directly within the terminal, responding to natural language commands. This assistant is designed to streamline various programming tasks, making your coding experience more efficient and intuitive.

Features

• Search the Internet: Quickly find coding solutions and references.
• Download Files: Automatically handle file downloads as needed.
• Execute Shell Commands: Run commands directly from the terminal.
• File Management: Read, write, and delete files with ease.
• Memory Management: Effectively manage internal memory for optimal performance.
• Vector Database Storage: Store information in a Weaviate vector database.
• Codebase Representation: Maintain a direct representation of your entire codebase for easy access and manipulation.

Setup Instructions

1. Install Requirements:

   pip install -r requirements.txt

2. Install:

    source install.sh

Environment Setup

• For OpenAI users:

  • Add your OpenAI API key to the .env file:

    OPENAI_API_KEY=your_api_key_here
    

• For Google/PaLM/VertexAI users:

  • Set up the Weaviate database using Docker:

    docker compose up -d

  • Add your Google credentials JSON file to the project folder.
  • Update the .env file with your Google credentials file path:

    GCRED_FILE=path_to_your_google_credentials.json
    

  • Add your PaLM key to .env

    PALM_APIKEY=your_api_key_here
    

Configure .botignore

• The .botignore file works similarly to .gitignore. Create this file in your project root directory and list the files or directories you want AutoCoder to ignore.

Assistant and Thread IDs

• After creating your assistant, save the assistant ID in the .env file:

ASSISTANT_ID=your_assistant_id_here

• Save the thread ID in the .env file for persistence:

THREAD_ID=your_thread_id_here