support

adk-python: Build AI Agents

Name: adk-python: Build AI Agents
Author: google

Get flexibility and control building AI agents with adk-python, a Python toolkit.

21,338 stars3,917 forksPythonGuide quality 9/10Updated 8/31/2026100% free · open source

What it does

The adk-python toolkit allows startup founders to build, evaluate, and deploy sophisticated AI agents with flexibility and control using Python.

When to use it

•You need to integrate AI decision-making into your application
•You want to evaluate and compare different AI agent architectures
•You require fine-grained control over the AI development and deployment process

Ready-to-paste prompt

python examples/run_example.py --agent=my_agent --env=my_env

Heads up: Make sure you have Python 3.8 or later installed, as adk-python is not compatible with earlier versions

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/adk-python && curl -fsSL https://workflowstacks.com/api/skills/adk-python/claude-skill -o ~/.claude/skills/adk-python/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How adk-python: Build AI Agents works

Codeflow

Free to inspect

Adk-python: Build AI Agents is a very large Python project (~239k lines across 1250 code files, plus 717 test files). Setup is light: installs like a normal app. Reading the code is optional. Last commit this month, Apache-2.0 license, has a test suite.

Size

Very large codebase

~239k lines · 1250 code files · days to read — use, don't read

Setup

One-command install

Installs like a normal app. Reading the code is optional.

Runs on

Python

No API keys detected

Python 98%Jupyter Notebook 2%

Where to start reading

1
README.md
Start here — what it does and how to install it
2
AGENTS.md
The instructions the AI actually follows
3
pyproject.toml
Dependencies and the commands it exposes
4
src/google/adk/__init__.py
Inside src/ — the main logic begins here

What's in each folder

src/Core code — the actual logic840 files

docs/Documentation46 files

contributing/Folder799 files

scripts/Helper scripts9 files

tests/Tests — proof it works808 files

.agents/Tool / agent settings62 files

.github/CI / automation (GitHub Actions)22 files

assets/Images & static assets3 files

READMEHas testsDocumentedCI checksApache-2.0 licenseUpdated this month