A2A_Example

# Kansai Dialect Agent Server This project implements an AI agent server that communicates in the Kansai dialect of Japanese. It utilizes the `a2a-server` framework, Google Agent Development Kit (`google-adk`), LiteLLM, and a Google Gemini model. Thanks a lot Chris Hay for his easy to understand explanation. This code based on his code base, please go check: https://github.com/chrishayuk/a2a-server ## Features * Responds to prompts in Kansai dialect (Kansai-ben). * Built using the `a2a-server` and `google-adk` libraries. * Uses LiteLLM to interact with the Gemini LLM. * Configurable via `agent.yaml`. ## Requirements * Python >= 3.13 * `uv` package manager (recommended) * A Google Gemini API Key ## Setup 1. **Clone the repository:** ```bash git clone <your-repository-url> cd test ``` 2. **Set up virtual environment and install dependencies:** ```bash # Create and activate virtual environment (using uv) uv init # Install dependencies uv add a2a-server uv add google-adk uv add litellm uv add python-dotenv ``` *Note: Dependencies are managed via `pyproject.toml` and locked in `uv.lock`.* 3. **Configure Environment Variables:** Copy the example environment file and add your API key: ```bash cp .env.example .env ``` Now, edit the `.env` file and replace `YOUR-GEMINI-API-KEY` with your actual Gemini API key. ## Running the Server Start the server using: ```bash uv run -m a2a_server --config agent.yaml ``` The server will start, typically on `http://localhost:8000` (as configured in `agent.yaml`). ## Using CLI After starting the server, use: ```bash uvx a2a-cli ``` ## Configuration The agent's behavior, server port, and metadata are configured in `agent.yaml`. ## Dependencies * `a2a-server`: Framework for building agent servers. * `google-adk`: Google Agent Development Kit. * `litellm`: Library for interacting with various LLMs. * `python-dotenv`: For loading environment variables from a `.env` file.