https://github.com/sushant1827/google_adk
This repository contains examples for learning Google's Agent Development Kit (ADK), a powerful framework for building LLM-powered agents.
https://github.com/sushant1827/google_adk
agentic-workflow google-adk litellm multi-agent-system persistent-storage pydantic session-storage structured-output
Last synced: 3 months ago
JSON representation
This repository contains examples for learning Google's Agent Development Kit (ADK), a powerful framework for building LLM-powered agents.
- Host: GitHub
- URL: https://github.com/sushant1827/google_adk
- Owner: sushant1827
- Created: 2025-05-04T08:19:35.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2025-05-27T10:44:55.000Z (5 months ago)
- Last Synced: 2025-05-27T11:40:42.619Z (5 months ago)
- Topics: agentic-workflow, google-adk, litellm, multi-agent-system, persistent-storage, pydantic, session-storage, structured-output
- Language: Python
- Homepage:
- Size: 105 KB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Agent Development Kit (ADK) Crash Course
This repository contains examples for learning Google's Agent Development Kit (ADK), a powerful framework for building LLM-powered agents.
## Getting Started
### Setup Environment
You only need to create one virtual environment for all examples in this course. Follow these steps to set it up:
```bash
# Create virtual environment in the root directory
python -m venv .venv
# Activate (each new terminal)
# macOS/Linux:
source .venv/bin/activate
# Windows CMD:
.venv\Scripts\activate.bat
# Windows PowerShell:
.venv\Scripts\Activate.ps1
# Install dependencies
pip install -r requirements.txt
```
Once set up, this single environment will work for all examples in the repository.
### Setting Up API Keys
1. Create an account in Google Cloud https://cloud.google.com/?hl=en
2. Create a new project
3. Go to https://aistudio.google.com/apikey
4. Create an API key
5. Assign key to the project
6. Connect to a billing account
Each example folder contains a `.env.example` file. For each project you want to run:
1. Navigate to the example folder
2. Rename `.env.example` to `.env`
3. Open the `.env` file and replace the placeholder with your API key:
```
GOOGLE_API_KEY=your_api_key_here
```
You'll need to repeat this for each example project you want to run.
## Examples Overview
Here's what you can learn from each example folder:
### 1. Basic Agent
Introduction to the simplest form of ADK agents. Learn how to create a basic agent that can respond to user queries.
### 2. Tool Agent
Learn how to enhance agents with tools that allow them to perform actions beyond just generating text.
### 3. LiteLLM Agent
Example of using LiteLLM to abstract away LLM provider details and easily switch between different models.
### 4. Structured Outputs
Learn how to use Pydantic models with `output_schema` to ensure consistent, structured responses from your agents.
### 5. Sessions and State
Understand how to maintain state and memory across multiple interactions using sessions.
### 6. Persistent Storage
Learn techniques for storing agent data persistently across sessions and application restarts.
### 7. Multi-Agent
See how to orchestrate multiple specialized agents working together to solve complex tasks.
### 8. Stateful Multi-Agent
Build agents that maintain and update state throughout complex multi-turn conversations.
### 9. Callbacks
Implement event callbacks to monitor and respond to agent behaviors in real-time.
### 10. Sequential Agent
Create pipeline workflows where agents operate in a defined sequence to process information.
### 11. Parallel Agent
Leverage concurrent operations with parallel agents for improved efficiency and performance.
### 12. Loop Agent
Build sophisticated agents that can iteratively refine their outputs through feedback loops.
## Official Documentation
For more detailed information, check out the official ADK documentation:
- https://google.github.io/adk-docs/get-started/quickstart
## Support
Need help or run into issues? Join our free AI Developer Accelerator community on Skool:
- [AI Developer Accelerator Community](https://www.skool.com/ai-developer-accelerator/about)
In the community you'll find:
- Weekly coaching and support calls
- Early access to code from YouTube projects
- A network of AI developers of all skill levels ready to help
- Behind-the-scenes looks at how these apps are built
## Error handling
### 1. If you're seeing a blank web page after running adk web, here’s something that might help.
https://github.com/google/adk-python/issues/114
Adding the below to fast_api.py (.venv\Lib\site-packages\google\adk\cli) fixed the blank issue for me
import mimetypes
mimetypes.add_type("text/javascript", ".js", True)
Don't forget to clear your browser cache after making above changes.