Learning Google ADK

Learning Agentic AI Development with Google ADK Step by Step.

This is my code repository, which I have created by following Brandon Hancock's Agent Development Kit (ADK) Masterclass on Youtube..

There are some changes that I have made from the code that Brandon has so graciously made available for free on his Github repo for the video. For instance, I use uv for environment management instead of Python and pip as used by Brandon.

How did I setup my environment

I am using uv, which is an extremely fast Python package & project manager. You should first install uv for your respective OS. Installation instructions can be found here. Once you have uv installed on your machine, follow the steps below:

1. Create a root folder to hold all the code for this video course (e.g. ~/home/learning_google_adk). This can be created anywhere on your hard disk - I'll refer to this folder as $CODE_ROOT henceforth.
2. Open a terminal window (or command shell on Windows) and cd $CODE_ROOT folder.
3. Created a requirements.txt file with the entries as shown here
4. Created a local Python environment by running uv init . [--python 3.12] in the $CODE_ROOT folder. The --python 3.12 is optional and used to specify a specific Python version to use. If omitted, it will default to latest Python version available.
5. Activated the just created environment by running (on Windows: .venv\Script\activate.bat; on Linux/Mac: source .venv/bin/activate). You should see your command prompt change to reflect the new Python environment is in use.
6. Then run uv pip install -r requirements.txt or uv add -r requirements.txt from the $CODE_ROOT folder. This will install all the required packaged.

When you are cloning this Github Repo

Firstly, fell free to do so. I am sharing code for learning purpose only with no fit-for-purpose guarantees!

You should first install uv for your respective OS. Installation instructions can be found here. Once you have uv installed on your machine, follow the steps below.

Assume you are cloning this repo to c:\code folder on Windows or ~/code folder on Linux or a Mac. The git clone ... command will create a ~/code/learning_google_adk subfolder (on Linux/Mac) or a c:\code\learning_google_adk (on a Windows machine)

Part A: setting up your local Python Environment

1. cd ~/code/learning_google_adk (Linux/Mac) or cd c:\code\learning_google_adk (Windows) folder.
2. Run uv sync to duplicate the environment that was created by me - thats it!

Part B: setting up API key for use with Google's Gemini models

<< TODO >>

Once you have the environment ready, you should be able to follow Brandon's YouTube video and learn Google ADK.