Local Agent

An edge agent framework built in pure Python.

• Connects to local language model servers: AnythingLLM, LM Studio, Nexa, and Ollama
• Easily extensible with custom tools

Quick Start

1. Refer to the Setup Guide to configure the application.
2. Create a virtual environment and install Python dependencies:

   python3 -m venv venv
   
   # Windows PowerShell
   .\venv\Scripts\Activate.ps1
   
   # Mac/Linux
   source venv/bin/activate
   
   pip install openai httpx pyyaml requests pytest

3. Run the tests to verify the configuration

   # Windows PowerShell
    .\scripts\run_tests.ps1 -c -l
   
   # Mac/Linux
   ./scripts/run_tests.sh -c -l
   

4. Configure your local LLM server settings in config.yaml.
5. Run your chosen model server (e.g., AnythingLLM, LM Studio, or Nexa).
6. Run the local agent

   python main.py

Contributing

I welcome contributions! Please follow these steps:

1. Fork the repository.
2. Create a new branch for your feature or bug fix.
3. Make your changes, commit them, and push to your branch.
4. Create a pull request explaining your changes.

Special thanks to these community contributors:

• Suraj Ranganath

License

This project is licensed under the MIT License. See the LICENSE file for details.