My ADK Agents

Initial Setup

1. Set up Environment Variables: The project root contains a .env.sample file. Copy this file to .env within the same directory and fill in the required values (like API keys or tokens).

   cp .env.sample .env
   # Then edit .env

2. Ensure uv is installed: Follow the installation instructions at https://docs.astral.sh/uv/install/.

3. (Optional) Create the Virtual Environment: While uv run can create the environment automatically, you can create it explicitly if preferred:

   uv venv

   You generally do not need to manually activate the environment or run uv sync if you use uv run for executing commands.

Running the Agents

Use uv run to execute the adk web command. uv will automatically ensure the project environment (.venv) exists and is synced with the uv.lock file before running the command.

uv run adk web

The web interface will be available at localhost:8000.

Running with Docker Compose

1. Copy the sample environment file and edit it as needed:

   cp .env.sample .env
   # Edit .env to fill in required values

2. Build and start the service:

   docker compose up --build

   The web interface will be available at http://localhost:8000

3. To stop the service:

   docker compose down

Adding New Dependencies (for Developers)

This project uses pyproject.toml to define direct dependencies and uv.lock as the lockfile generated by uv for reproducible installs.

To add a new dependency:

1. Add the package using uv add: This command automatically updates both pyproject.toml and uv.lock.

   # Example: Add the 'requests' library
   uv add requests
   
   # Example: Add a package with a version constraint
   uv add "requests>=2.25"

2. Commit the changes: Commit both the updated pyproject.toml and uv.lock files to version control. This ensures that other developers and deployment environments use the exact same set of dependencies. Other developers will automatically get the new dependency the next time they use uv run or if they explicitly run uv sync.

Deploy

docker build -t cofacts/beta-ai .
docker push cofacts/beta-ai

And go to server to pull & run the image.