ADK Agents Playground

A collection of Google ADK sample agents demonstrating common patterns, including tool-augmented agents, function tools, LangChain integration, sequential workflows, parallel execution, and loop-based refinement.

Getting started

1. Create a Python environment

python3 -m venv .venv
source .venv/bin/activate

2. Install required packages

This repository uses Google ADK and related Python tooling. Install dependencies as needed for your environment.

pip install google-generativeai langchain langchain-community requests

If your project already has a specific dependency file, install from that instead.

3. Configure credentials

Make sure your Google ADK credentials and any provider-specific environment variables are set in your shell or .env files within each example folder.

Common variables may include:

• GOOGLE_API_KEY

Running an example

Each folder exposes a root_agent object, which is the example agent ready to execute. Use your ADK runtime or application code to load and run the sample agent.

Example usage pattern:

from loop_agent.agent import root_agent

# Use your ADK runtime to execute the root_agent with the expected inputs.
# For example, pass a prompt or data payload to the agent runtime.

print(root_agent)

The exact execution API depends on your ADK setup and runtime.

How to extend

1. Copy an existing example folder.
2. Build a new agent.py with a root_agent definition.
3. Add any custom tools, new workflows, or new model instructions.
4. Update this README with your new agent example.

Notes

• Each example is intentionally small and focused on one ADK pattern.
• root_agent is the primary entrypoint for each sample.
• Use .env files inside example folders to keep configuration local and separate.

Contributing

Contributions are welcome. Add new agents, improve instructions, or add sample prompts and workflows.

---

Built for the ADK agents playground with a practical focus on reusable patterns, tool integration, and workflow composition.